Xilinx Xcell Journal 36 - complete issue (Q2 00)
Carlis Collins [email protected]
BOARD OF ADVISORS
Pronto Creative www.prontocreative.com
2100 Logic Drive
San Jose, CA 95124-3450
©2000 Xilinx Inc.
All rights reserved.
Xcell is published quarterly. XILINX, the
Xilinx logo, and CoolRunner are registered trademarks of Xilinx, Inc. Virtex,
LogiCORE, Spartan, SpartanXL, Alliance
Series, Foundation Series, CORE Generator,
Checkpoint Verification, TimeSpecs, Smart
IP, QPRO, SelectI /O, SelectI/O+, True Dual-
Port, WebFITTER, WebPACK, Select RAM,
BlockRam, Xilinx Online, and all XC-prefix products are trademarks, and The
Programmable Logic Company is a service mark of Xilinx, Inc. Other brand or product names are trademarks or registered trademarks of their respective owners.
The articles, information, and other materials included in this issue are provided solely for the convenience of our readers.
Xilinx makes no warranties, express, implied, statutory, or otherwise, and accepts no liability with respect to any such articles, information, or other materials or their use, and any use thereof is solely at the risk of the user. Any person or entity using such information in any way releases and waives any claim it might have against Xilinx for any loss, damage, or expense caused thereby.
The Future of
W ithin the next five years, programmable logic will be the key technology for developing next generation products; within the next ten years, programmable logic devices (PLDs) will be used in virtually every electronic product on the market.
These industry predictions are based on a number of recent developments that have forever changed the way electronic systems are designed.
The old advantages of custom ASICs are quickly being overcome by the new advantages of programmable logic. Here are some of the reasons why: designs, either those created in-house or by third party suppliers. There is already a wealth of IP, and many new designs are being introduced every day. IP significantly reduces your development time and your risk.
Highly Efficient Development
Tools - Speed is critical. Not only do you want the fastest design you can create, but you want to complete that design with the least possible risk and effort. The currently available development tools from Xilinx allow you to quickly develop your designs in many different ways, with multiple developers located in different places. These tools are fast, efficient, and easy to use; and they are constantly improving.
Very High Density, System-level
Devices - We now produce high performance FPGAs with advanced features and up to 3.2 million system gates. Plus, we have just announced our next generation 10-million gate architecture. With devices this powerful and this dense you are limited only by your imagination.
Very Low Power, Low-cost
Devices - Many new designs require battery operation and low cost. Our Fast
Zero Power™ CPLDs are perfect for cell phones, PDAs and other power sensitive applications. Our Spartan FPGAs already give you 100K gates for less than
$10.00, and higher density, lower cost
FPGAs are on the way. Power and price are no longer an issue with PLDs.
Intellectual Property - The fastest and surest way to get your PLD-based product to market is to use proven
Field Upgradeability - With the current Xilinx device and software technologies you can create designs that are easily changed, in the field, over any network, right at your customers’ premises. The possibilities are enormous, and the advantages are overwhelming; this cannot be duplicated with any other technology. Now, just like software, your hardware can easily change to meet the demands of the marketplace.
With programmable logic you not only get the fastest and easiest way to create next-generation systems, you also get significant advantages that cannot be offered by fixed logic ASICs; there is no better way to create value.
For A FREE
To The Xcell
E-mail your request to: [email protected], please be sure to include:
1. Your full name and mailing address
2. Your title
3. The name of your company
4. Your e-mail address
5. Is this new subscription or a subscription renewal?
Inside This Issue:
View from the Top
View from the Top
Field Upgradable Systems
by Wim Roelandts, President and CEO, Xilinx
O ne of the most exciting developments in our industry is the increasing shift toward using networks, including the Internet, to remotely upgrade the digital equipment that is already installed at a customer’s location, anywhere in the world, and beyond.
Updating software remotely, with new enhancements and bug fixes, is a common practice. However, remotely updating hardware may appear to be more challenging because hardware is typically a fixed entity that is updated by manually replacing circuit boards. Now, with our current FPGA families and the
Xilinx Online software technologies, doing automatic, remote hardware upgrades is not only a reality, it is becoming a necessity in many new applications.
Researchers have been investigating this concept for years, and a few forward looking companies have already deployed FPGA-based remotely reconfigurable products. IBM, for example, currently markets an ATM switch whose FPGA-based logic can be changed over the network to bring it into accord with the latest changes in the ATM standard. A number of Xilinx customers, including large communications companies building the next generation of WCDMA wireless systems, are very interested in the idea, and a few are well along the way with major designs efforts.
The advantages of this new possibility are enormous. For example, the ability to remotely update hardware with new features or the latest bug fix can accelerate your time-to-market, extend the useful life of existing systems, and significantly cut production, maintenance, and support costs. Many of today’s systems already come with some form of built-in communications or microprocessor interface, making the addition of remote field update capability a simple matter. If you consider doing remote updates during the initial specification and design process, your systems can easily reap all the benefits of being updated remotely.
Remote upgradability significantly increases the useful lifetime of a system. The ability to add new hardware features and fix existing ones without sending a technician out to the field can add up to considerable maintenance and support savings over the entire life of the system. Imagine the implications for satellite-based communications equipment.
Remotely upgradeable systems can also provide new revenue prospects. After the initial product is released, you can develop new hardware features then sell and distribute those features to existing customers just as software developers do today. Or a standard “off-the-shelf” application can be developed so features can be swapped in and out depending on what the end-user purchases or needs.
Any system that has some type of connectivity to the “outside world” could potentially benefit from being designed to support field updates. Typical products include network appliances, set-top boxes, security systems, network equipment, cellular base stations, and satellite communications systems. Other likely applications are HDTV, video and image processing, encryption, military communications, surveillance, radar, and sonar.
Clearly, electronic equipment manufacturers who begin to think about the benefits of remote hardware upgrades today will be the ones who lead their markets in the not-to-distant future.
Hardware Systems Using
Enabling Technology from
A complete system for managing and remotely upgrading your FPGA designs in the field.
by Greg Brown, Xilinx Software Product Manager for Internet Applications, [email protected].com
and by Mike Akers, GoAhead Software FieldUpgrader Product Manager, [email protected]
F or many years designers have been using
Xilinx programmable logic devices (PLDs) to reap the benefits of fast time-to-market.
Now, a second major advantage is made possible through the partnership of GoAhead
Software and Xilinx—the ability to upgrade your designs via a network after they have been deployed in the field. This can result in dramatic field service cost savings and it helps “futureproof” your product.
The Solution for Field Upgrading
The new Xilinx and GoAhead Software solution combines the GoAhead commercial software
(GoAhead FieldUpgrader™) with the Xilinx
Internet Reconfigurable Logic methodology
(IRL™) to provide the enabling technology to help you create and manage field upgradable systems. This technology provides the backbone for network-based system designs as well as the delivery mechanisms to successfully deploy upgradable products.
The GoAhead FieldUpgrader software consists of three parts:
• GoAhead DeviceStudio™ - a development environment used to create the UpgradeAgent software.
• GoAhead UpgradeAgent™ - the software that is embedded in the target device. Each agent is unique to the particular operating system that is resident on the target system.
The target system containing the
UpgradeAgent polls the UpgradeServer at regular pre-determined intervals to look for
FPGA upgrades. When an upgrade is available, the device downloads the upgrade to a predetermined storage location. This process is illustrated in Figure 1.
• GoAhead UpgradeServer™ - used to create and publish upgrades and is usually located on a server at the manufacturer’s site. When the need for an FPGA upgrade arises, the manufacturer publishes the upgrade with
Device initiated request
Upgrade request response
Request for payload
Figure 1 - The remote upgrade process.
Data Security and Integrity
To upgrade devices over a network, it is critical that the devices only accept upgrades from the authorized, designated server. It is also important to ensure that the upgrade payload is not modified or corrupted. GoAhead FieldUpgrader and the UpgradeServer use a variety of techniques to authenticate the upgrade and prevent modifications along the way.
GoAhead FieldUpgrader uses the Digital
Signature Standard (DSS), which specifies a
Digital Signature Algorithm (DSA). The DSA provides the capability to generate and verify signatures using public and private keys. In addition, the device-initiated approach provides built-in security; the device does not accept externally initiated upgrades. The location and port of the upgrade server is configured into the target device and the device will only allow upgrades from this designated upgrade server. This method ensures that the upgrade payload received at the target device matches the original upgrade payload published at the upgrade server and guarantees that the message was not modified en-route to the device.
The payload of the upgrade is broken into smaller chunks to make the transfer more efficient and re-startable. GoAhead FieldUpgrader calculates a 32-bit CRC checksum for each of these individual chunks. If the target device detects a checksum error, it discards that chunk and re-fetches it from the upgrade server. Once all of the chunks have been downloaded, the device reconstructs the payload and performs the higher-level data integrity check.
Data integrity is also checked during the actual programming of the Xilinx FPGAs; there is a CRC checksum built into the programming step.
The UpgradeAgent is fault tolerant during the transmission of the upgrade. The Agent knows the contents of the payload based on the manifest list and will re-initiate the upgrade to fetch any missing chunks if the connection with the server is lost.
The application of the upgrade to the Xilinx device can also be made fault tolerant in the system architecture, which is shown in Figure 4.
The concept uses redundant non-volatile storage areas—one to hold the current, known good system, and the other to hold the upgrade. The system can always fall back to the known good configuration if there are any issues associated with upgrading the system.
Flexibility By Design - Pull or Push?
There are some applications that may require a
“push” method of upgrading. The term “push”
entails the directing and initializing of upgrades from a central server. There are cases or management decisions that may dictate this alternative to providing upgrades.
The FieldUpgrader software primarily uses a device-initiated or “pull” methodology. However, it was developed to be flexible as well, and can be enhanced to allow a “push” methodology. At a pre-determined time, the UpgradeAgent may simply open a communications port to listen for
“hello” messages that are broadcasted or multicasted from the UpgradeServer. If the
UpgradeAgent responds after verification, the
UpgradeServer sends a request packet to the system to begin the upgrade process. The
UpgradeAgent then makes the request for an upgrade from the secure UpgradeServer. This methodology solves the problems associated with firewalls and security. The requests may pass through firewalls and the device port remains open for only the specified upgrade period.
GoAhead FieldUpgrader supports the following platforms:
• GoAhead UpgradeServer, GoAhead
Windows NT4.0, Linux 2.2, Microsoft
Internet Explorer 4.x or greater, Netscape
4.x or greater
• GoAhead UpgradeAgent:
Wind River VxWorks 5.3.1 or greater (x86,
PowerPC, ARM, MIPS), Windows 95/98/NT
(x86, PowerPC), Windows CE (x86, Hitachi
SH) or Linux 2.2 (x86)
(Windows NT 4.0)
Payload Package upgrade .js
The key requirements of the UpgradeAgent are:
• TCP/IP stack and connection.
• Real-time clock.
• File system (optional for VxWorks).
• Provision of the main software module’s C source code (main.c) and a linkable library to enable the integration of the UpgradeAgent into custom applications.
- Control the upgrade process
- Load and call C functions to extend functionality
- Send additional request data to GoAhead
- Access GoAhead UpgradeAgent environment variables
The demonstration design utilizes the GoAhead
FieldUpgrader to upgrade a Xilinx XCS05 Spartan
FPGA, which is connected to a popular embedded system—a single-board computer running
Wind River’s VxWorks ® real-time operating system. This type of system exists in many applications and environments from industrial control stations to remote monitoring systems. Figure 2 shows a diagram of the system.
Field Upgrader Agent
Dynamically loads xiprog.o
Programs XCS05 with test bit
Xilinx Demo Board
Figure 2 - Demonstration system.
This demonstration system includes the following components:
Processor - The processor is a single-board,
Intel 486-based system. The board has 2MB of
RAM, integrated VGA controller, an IDE controller, and a parallel interface. A hard disk is used for non-volatile storage other than the
Xilinx demonstration board - This board has two devices on it: an XC3020A and a Spartan
XS05. The Spartan XS05 is a 5000 system gate
FPGA that is used in high volume, low-cost applications.
Xilinx Parallel Cable III - This cable runs between the single board computer and the
Xilinx demonstration board. This cable is used to provide the communication between the computer system and the Xilinx FPGA.
Wind River VxWorks - The VxWorks ® Real-
Time Operating System (RTOS) from Wind River
Systems is the most widely used RTOS for embedded systems.
The GoAhead UpgradeAgent - This is a realtime embedded application that uses VxWorks’ dynamic loading capability to load and execute other real-time embedded applications. In the demonstration design, this feature of VxWorks is used to dynamically load the object code that programs the Xilinx FPGA. This was strictly an architectural choice. An alternative and equally valid approach is to have the object code that programs the Xilinx device be resident on the
VxWorks system and not include it as part of the upgrade process, or use a combination of the two approaches.
The UpgradeAgent is configured to poll the
UpgradeServer once each minute to check for upgrades. When one exists, it downloads the upgrade payload and executes the instructions included in the payload. For demonstration purposes, two payloads are published on the
UpgradeServer: one to upgrade the bitstream and one to downgrade to the original bitstream.
This enables the system to toggle between the two FPGA configurations.
UpgradeServer - The UpgradeServer is running on a laptop PC using Microsoft NT4.0. It waits until the UpgradeAgent on the VxWorks system contacts it and requests an upgrade. If there is an available upgrade, the server sends the upgrade payload to the VxWorks system across the TCP/IP network.
Demonstration Upgrade Process
The upgrade process used in the demonstration design is illustrated by the flow diagram in
Demonstration System Upgrade Process
VxWorks programs the
XS05 through parallel port using bit file on disk
Upgrade Agent waits on event schedule
Upgrade Agent build
Upgrade Agent sends request to Upgrade Server
Upgrade Agent unarchives a payload
(Xiprog.o, upgrade.js, test.bit)
Upgrade Agent verifies payload
Upgrade Agent rebuilds archive file
Manifest is sent to
Upgrade Agent fetches upgrade chunk to disk
Figure 3 - Demonstration system upgrade process.
Building a Commercial System
There are several potential field upgradable architectures to which the concepts presented here apply. However, only one is described here—the single board computer running an
RTOS with programmable logic, as shown in
In Figure 4, the processor must be one supported by the GoAhead UpgradeAgent. This can
Storage 2 is ready to contain the next upgrade.
Possible selections for the non-volatile memory are EEPROMS, FLASH memory cards, or the
Xilinx XC18V00 in-system programmable configuration PROMs. (See page 62.)
Figure 4 - Basic system architecture
be either an Intel X86 (or compatible), PowerPC,
ARM, or MIPS processor running VxWorks as the
RTOS (or a supported Microsoft Windows or
Linux configuration as previously listed).
The non-volatile storage can be divided into three functions:
• Processor Program Memory (Non-Volatile
Storage 1) - contains firmware for the microprocessor.
• FPGA Configuration Storage (Non-Volatile
Storage 2) - contains configuration data for the FPGA. The system can be designed so that this storage area contains the initial configuration data.
• FPGA Configuration Storage (Non-Volatile
Storage 3) - contains configuration data for the FPGA. The system can be designed so that this storage area contains the first upgrade configuration data.
The system should be designed to always have a last known-good configuration for the
FPGA. This is handled by the redundancy of the
Non-Volatile Storage 2 and Non-Volatile Storage
3. In a usual sequence of events, the FPGA will be initially configured from Non-Volatile Storage
2. When an upgrade is requested, it is downloaded into Non-Volatile Storage 3. The FPGA is then reconfigured from this location. The program controlling the programming of the FPGA then switches the functions of these two storage areas as Non-Volatile Storage 3 is now the known-good configuration and Non-Volatile
There are two basic software components that need to be written for the target processor architecture:
• Firmware to copy the Xilinx programming file to non-volatile storage.
• Firmware to program the FPGA.
Another approach is to use the Java programming language either with the firmware or as a complete application (Xilinx supplies a Java
API for the Boundary Scan configuration mode).
A Java application can therefore be written (and called by the UpgradeAgent) to program the
FPGA. On Windows platforms, there are several
Java run-time engines (the virtual machine) available. On a VxWorks platform, the application would need to be developed using Personal
JWorks™ from Wind River.
The purpose of the GoAhead and Xilinx partnership is to provide the enabling technology and solutions for a variety of field upgradable systems. GoAhead FieldUpgrader provides the management, deployment, communications, and security backbone, while Xilinx provides reconfigurable hardware devices and programming technologies. Together, you have the building blocks needed to enable the rapid development and deployment of a new generation of costsaving, life-extending, field-upgradable systems.
For more information regarding how the Spartan-II family addresses traditional ASIC and ASSP designs, please see the article on page 49.
For more information see: www.xilinx.com/xilinxonline/ partners/goaheadhome.htm
Technology Update - CoolRunner
Fast Zero Power
CoolRunner CPLDs require very little power yet operate at very high speeds—here’s how.
How CoolRunner CPLDs Minimize System Current Demand
by Ron Cline, Director of CoolRunner Product Development, Xilinx, Inc., [email protected]
I n traditional CPLD architectures, including the XC9500 series, the circuits that propagate logic-level transitions in the product-term array are derived from the original (old) bipolar
PLD designs. These older designs use sense amplifiers at the end of each bit line in the product-term array to achieve fast propagation delays. Because CPLD product terms cannot be decoded (as with memory locations in an
EPROM, for example) there must be a sense amplifier for each individual product-term. These sense amplifiers must be continuously operational, and will draw continuous supply current even when not switching.
MUXs are programmed using RAM-based configuration bits.
The full CMOS AND gate shown in Figure 1 has a delay of under 0.5 nsec, including the delay of the input MUXs.
The CoolRunner Design Technique
The Xilinx CoolRunner design uses an innovative method for implementing the product-term array. Rather than employing sense amplifiers (a bipolar-style circuit), CoolRunner CPLDs use true
CMOS circuitry. In the CoolRunner Fast Zero
Power (FZPTM ) approach represented in Figure
1, the AND gates in the product-term array are implemented using configurable multiplexers
(MUXs) attached to the inputs of normal CMOS
NAND gates. Each MUX selects an input, it’s complement, or Vcc (don’t care state.) These
Figure 1 - CoolRunner Fast Zero Power AND gate.
Output = I
A four input ‘AND’ function demonstrates how the product term is implemented with full-CMOS gates in the CoolRunner CPLDs.
Figure 2 - Representation of a 4-input product-term using the CoolRunner FZP design technique.
Output = I
The previous example, expanded to implement an eight-input AND function.
Figure 3 - Expanding the number of inputs.
Wider AND gates are built using a deMorgan tree, as shown in Figure 2. Doubling the input width simply requires the replacement of the inverter in Figure 1 with a 2-input NOR gate.
This increases the total delay by less than 0.1
This design technique can be extended for wider widths, as shown in Figure 3 which shows a re-doubling to eight inputs, at an additional delay penalty of less than 0.2 nsec for the NAND gate plus inverter. The inverter can be replaced with a two-input NOR gate to enable a productterm width of 16 inputs (at an incremental delay penalty of 0.1 nsec.).
As you can see, the delay penalty per additional input actually decreases as the number of inputs into each product-term increases. This is in contrast to the traditional sense amplifier approach, where the delay increases linearly as product-term input width is increased.
The gate tree implementation distributes the capacitance at each product-term, so this capacitance is no longer lumped on a single node.
Furthermore, the switching current behaves in a manner similar to that of random logic in a gate array; the static current for each gate is smallabout 1 picoamp (pA). The total instantaneous dynamic current is also small, because only the gates in one path of the tree can switch, and these gates switch in succession rather than all at once.
The advantages of CoolRunner FZP CPLDs are numerous. Total standby current is under 100 microamps—at least 1000 times less than that exhibited by CPLDs using sense amplifiers. Total dynamic power is also decreased, relative to existing CPLDs, by as much as 70% for a device whose logic is fully populated with 16-bit counters operating at 50 MHz. Best of all, these power savings are realized with little impact on performance and with no cumbersome powerdown circuitry.
Because power consumption is so low, chipscale packaging options, once considered impossible due to thermal limits, are now offered. As a result, FZP technology is a key technology for
CPLDs because it enables very low-power, highperformance applications.
New Products - Development Tools
The CoolRunner XPLA3 Development Kit, from Xilinx and Insight
Electronics, is the perfect way to get started using the most efficient, low power CPLD on the market.
by John Hubbard, Application Engineer, Xilinx, Inc., [email protected]
X ilinx and Insight Electronics have jointly developed a low power demonstration board that allows you to easily experiment with the new CoolRunner XPLA3 architecture, using In System Programming (ISP) to configure the device. With this board, and the free
Xilinx PC-ISP3 Programmer software, you can immediately begin configuring a CoolRunner
XPLA3 device with your custom design, directly from your PC. If you need a little help getting started, you can get the XPLA3 Demo Board
User’s Manual or the VHDL/Verilog tutorial from the Xilinx Application Note website.
Figure 1 - The XPLA3
down mode and is the only CPLD family delivering both high performance and ultra low power consumption.
The XPLA3 CoolRunner family (Figure 1) has the lowest power consumption of any CPLD family on the market. The entire CoolRunner XPLA3
Development Board, with its pre-programmed
“CoolrunnEr” scrolling marquee pattern, draws only 60 µA. Since the board contains other components that also draw current, such as the onboard low power oscillator and LCD, the
CoolRunner CPLD actually draws less than
60 µA. Because the CoolRunner product line requires so little power, it does not need a power
The development board, shown in Figure 2, is designed to demonstrate the low power requirements and the advanced features of the
CoolRunner architecture. The board uses the
Xilinx XCR3256XL, a 256-macrocell device in a
TQ144 package, which can be programmed via the JTAG ISP port using the provided Xilinx
Figure 2 - The CoolRunner XPLA3 Development Kit uses only 60 µa (powered here by six grapefruit).
Parallel Download Cable III; any JEDEC file that targets the XCR3256XL can be downloaded.
To power the board you can use either:
• A 3.3V user regulated input.
• A 10.0V maximum input, regulated to 3.3V
using the on-board regulator.
• A 6.0V AC adapter input regulated to 3.3V
using the on-board regulator. This AC adapter is included with the CoolRunner XPLA3
Two clock sources are available to the
• The on-board, low power 32.768 kHz oscillator.
• Any external clock source with an impedance of 50 ohms (to match the 50 ohm on-board trace impedance).
A prototyping area is also provided, and all device I/Os are available for your prototype either via the user access headers that surround the CoolRunner device or via the 20 throughhole connections routed to specific I/Os. The two-digit seven-segment LCD (shown in the lower left corner of the board in Figure 2) is connected to specific I/Os so you can quickly and visually verify your design; it can be disconnected from the device if you choose. Control pins are also included so you can easily implement the Xilinx Watch Tutorial which is available for download from http://www.xilinx.com/ apps/epld.htm#tutorials.
You can quickly and easily become an XPLA3
CoolRunner guru by using the CoolRunner
XPLA3 Development Kit; and it’s available today from Insight Electronics.
Order the CoolRunner XPLA3 Development Kit from the Insight website at: http://www.insight-electronics.com/xcellence/scalable/kit/.
The part number is DS-XPLA3-PAK and the price is $95.00.
Download the free PC-ISP3 Programmer software from the Xilinx
WebPACK website at http://www.xilinx.com/products/software/ webpowered.htm.
Get the XPLA3 Demo Board User’s Manual or the VHDL/Verilog tutorial from the Xilinx Application Note website at: http://www.xilinx.com/apps/appsweb.htm.
Visit the XPLA3 CoolRunner website, for the latest CoolRunner information, at: http://www.xilinx.com/products/xpla3.htm.
Products CoolRunner CPLDs
Implementing a 16B/20B
Encoder/Decoder in a
Here’s an overview of a complete design that you can download from the Web.
by Jennifer Jenkins, Applications Engineer, Xilinx Inc., [email protected]
T he 8B/10B data transmission scheme has become the standard for high-speed serial links today. This encoding scheme translates byte-wide data of random ones and zeros into a 10-bit serial data stream. The
8B/10B encoding rules create a DC balanced code that provides optimum coding efficiency, clock recovery, error detection, and suitability for ring or point-to-point topologies.
Parallel byte of data
6B Bit Control
4B Bit Control
10 binary lines to serializer
i d e a b c f g j h
The 16B/20B transmission scheme incorporates the idea of the 8B/10B transmission code by combining two 8B/10B modules side-by-side.
With 16B/20B encoding, a 16-bit word can be encoded and transmitted serially as shown in
Figure 2 shows the block diagram for each
8B/10B encoding module. The 8B/10B transmission code includes both data and control characters. Parity is monitored in each data byte sent and determines the encoded data. Each byte of
Serial Data Out
Serial Data In
Figure 1 - 16B/20B block diagram.
Figure 2 - 8B/10B encoder logic block diagram.
data is broken into 5B/6B and 3B/4B encoding functions. The parallel byte of data to encode
[0:7] is referenced A through H, respectively. The data is then transmitted serially with control characters i and j.
The 16B/20B encoder module or 20B/16B decoder module is targeted to a CoolRunner
XPLA3 CPLD. CoolRunner CPLDs not only provide the lowest power solution today, but entering a design is simple with the WebPOWERED software tools available. WebPACK™ and
WebFITTER™ allow for easy design entry as well as simulation and implementation. The
WebPOWERED software tools can be downloaded for free from the Xilinx website, or by going to www.xilinx.com/products/software/ webpowered.htm
More Information: For a complete 8B/10B code description and VHDL implementation, look at XAPP336 “Design of a 16B/20B
Encoder/Decoder using a CoolRunner CPLD” on the Xilinx website, www.xilinx.com or by contacting Xilinx Technical Support at
New Products - FPGAs
Over 1-Mbits of
Xilinx extends the Virtex family; new FPGAs use copper interconnect technology for optimized performance.
by Rob Schreck, Product Manager, Xilinx, [email protected]
M any high-performance applications in networking and video processing require large amounts of on-chip RAM. To meet this demand, Xilinx recently announced the Virtex™-E
Extended Memory (Virtex-EM) FPGA family that is the first in the industry to provide over one megabit of
True Dual PortTM Block RAM and is also the first to offer copper interconnect for optimized performance.
Xilinx primarily developed the Virtex-EM family to address the application requirements of our large networking customers. The unique combination of block
RAM and logic, along with copper interconnect, allows increased data bandwidth and greater integration; this gives networking companies a very efficient and reliable platform for their 160 Gbps switch fabrics.
The Virtex-EM family builds on the highly successful
Virtex architecture and includes the following:
• Two devices: XCV405E and XCV812E.
• Over 1-Mbit of True Dual-Port RAM (XCV812E).
• Leading edge 0.18 micron, 6-layer metal silicon process.
• New copper interconnect used for top two layers:
- Top layer for uniform on-chip power distribution
The Virtex-EM family also addresses high-end video processing, which uses DSP engines to improve image quality and implement video compression and decompression algorithms. By using the large block
RAM capacity, both the line buffer memory and either a video processing engine or a DSP function can be accommodated on the same Virtex-EM device, which greatly enhances the overall system bandwidth.
- Second layer for clock signals - minimizes clock skew
• Over 311 Mbps single-ended I/O performance.
• 622 Mbps differential I/O performance.
Availability and Pricing
Samples of the XCV812E are available now, and the
XCV405E will be sampled in June. Both devices are expected to be in full production in third quarter this year. The XCV812E and the XCV405E devices are
• Complete hierarchy of memory resources.
priced at $235 and $101, respectively for 50,000 units
• Eight DLLs for over 311 MHz clock management.
in Q4 2000.
• Direct interface to high-performance external memory.
For more information on the Virtex-EM FPGA family see: http://www.xilinx.com/products/ virtex/ss_virem.htm.
Virtex FPGAs are used in communications, networking, and video image processing applications, where
New Products - Software
Xilinx Software R&D delivers new version 3.1i software tools that empower you to maximize your productivity, while leveraging your creativity.
by Craig N. Willert, Software Marketing Manager, Xilinx, [email protected]
P rogrammable logic design has entered an era where device densities are measured in the millions of gates, and system performance is measured in hundreds of
MegaHertz. Given these new system complexities, the critical success factor in the creation of a design is your productivity. Version 3.1i of the
Xilinx development systems were created with your goals in mind-harnessing your creative engineering talent to the greatest extent possible.
Maximizing Productivity - Keeping Your
Creating your programmable logic design is probably only part of your job. You may also have to complete the board design, verify the design in the lab, or help create your product demonstration. We understand that Xilinx can contribute to your success by ensuring that the
Xilinx design flow is as productive as possible.
The 3.1i tools include the following improvements, making it easier than ever to finish your programmable logic design on time:
• Faster runtimes (2x).
• Faster clock Speeds (2x).
• Advanced HDL synthesis and optimization.
• Improved timing analysis and error navigation.
• Integrated logic analysis.
Xilinx development systems deliver run time improve-
Figure 1 - Runtime improvements over consecutive releases.
ments that once again cut place and route runtimes in half. This is the fourth consecutive software release with 50% runtimes reductions, which means that the Xilinx solution is completing designs 16X faster than 4 years ago, as shown in Figure 1. It is important to note that these runtime improvements are independent of computing platform, and achieved even when placing and routing today’s more complex devices. Improvements like these are why Xilinx is able to help keep you productive even when you are designing with a multi-million gate
Quality of results
To be most productive, you also need place and route tools that reliably converge on
2X your design’s timing requirements. Xilinx invented Timing
Driven Place and Route for programmable logic
in 1992, and through our continued algorithmic innovations, we remain the leader in delivering superior quality results. In the 3.1i release, you can achieve 2x performance improvements when compared with designs created using the fastest speed grades that were available at the time of the previous 2.1i release. The industry’s most advanced timing driven place and route tools, coupled with fastest devices, allow you to meet even the most challenging timing requirements using Xilinx programmable logic.
PC with Chipscope
Target FPGA with ILA cores
Advanced HDL Synthesis and Optimization
Figure 2 - Integrated logic analysis.
Xilinx 3.1i Foundation Series products include a new Block Level Incremental Synthesis (BLIS) capability that enables you to rapidly converge on your design’s timing requirements. Synopsys has built this capability exclusively for Xilinx.
BLIS is based on the concept of focussing the optimization of HDL on a module by module basis, localizing the changes within a module, and therefore improving the efficiency of guided design. BLIS is yet another design methodology that Xilinx is delivering to help you complete your design more efficiently.
archical browsing capability. The new Timing
Analyzer also includes a powerful “what if” analysis feature that allows for immediate analysis of your design when targeting different device speed grades, or using different timing constraints. Furthermore, Xilinx has rewritten its timing analysis algorithms so it requires substantially less memory and processes designs far faster. The combination of these improvements means it is easier to quickly analyze the performance of your design.
Error Navigation and Timing Analysis
Integrated Logic Analysis
Having thorough and informative reports on the processing of your design also increases your efficiency. Furthermore, presenting this information in an intuitive and efficient format is also important; reviewing the reports is now easier than ever:
• Error Navigation - For warnings and errors that may arise during design compilation,
Xilinx provides a direct link to a wealth of
Solution Records that exist on our website
(www.support.xilinx.com). These Solution
Records provide answers to problems as simple as “incorrect pin connections” as well as more complicated problems like “how to obtain better clock rates using digital Delay
• Timing Reports - For information on your designs timing, the Xilinx Interactive Timing
Analyzer includes a new and improved hier-
Your job isn’t done until after your design has been incorporated into the production system as illustrated in Figure 2. To assist you in verifying your programmable logic device’s behavior within the system, Xilinx now provides
ChipScopeTM ILA, the Integrated Logic Analyzer.
With ChipScope ILA you can perform complex system analysis on the behavior of your design as it works in real-time, within the system. The
Integrated Logic Analyzer allows you to capture the waveforms of control signals, data buses, or any general logic within your Xilinx FPGA, and easily analyze their timing and functionality using ChipScope’s easy-to-use graphical user interface. Helping you detect problems in the functionality or timing of your design within your system is another example of how Xilinx helps support you in becoming successful. (See the
Emphasizing Creativity - How Good
Designs Become Great
Mainstream design flows enable you to enter your design’s functional and timing requirements and then “hand them off” to the layout tools to complete the job. Xilinx development systems are engineered to simplify the process of communicating your expert knowledge to our advanced design algorithms.
The 3.1i release includes new and improved versions of the following implementation tools:
• Xilinx Core Generator.
• Xilinx Constraints Editor.
• Xilinx High Level Floorplanner.
This powerful collection of tools will help make the process of creating your next programmable logic design the most efficient and productive yet.
By allowing you to easily reuse Intellectual
Property, built elsewhere within your organization or by a 3rd party provider, the Xilinx Core
Generator allows you to focus your creative energies on creating the unique aspects of your design, ensuring your products’ success in the marketplace. The 3.1i release of the Core
Generator delivers improvements in the graphical user interface, design flows, plus there is a host of new cores. See the IP Center http://www.xilinx.com/ipcenter/index.htm to learn about the Intellectual Property that is available for your use.
In the 3.1i release, the Xilinx
Floorplanner has been enhanced to deliver the benefits of true modular design. The 3.1i floorplanner may be used in conjunction with the Xilinx
Modular Design tools to enable the independent
Timing Driven layout of functional modules of a design. Another benefit of the new Floorplanner is the new pin layout capability, which makes it simple to assign your design’s I/O signals to the desired device pins.
Xilinx Constraints Editor
The 3.1i Xilinx Constraints Editor is another huge step forward in timing driven place and route efficiency. It simplifies the art of defining your circuits timing specifications, taking advantage of the industry’s most robust timing constraint language—TimeSpecs™. In 3.1i, the Constraints
Editor also supports the definition of timing constraints on a hierarchical basis, thus supporting modular design.
Furthering the art of logic design
The Xilinx 3.1i release has already been heralded as the industry’s best programmable logic design tool. In the future, you will enjoy even more productivity because the Xilinx Software R&D team is already hard at work inventing tomorrow’s break-through design techniques. While many of these inventions won’t be available for mainstream use until our next major software release, some of these advancements will be included within our regular quarterly software updates.
Please stay tuned for further announcements about your state-of-the-art programmable logic design system.
By focussing our resources on the challenges of productivity, Xilinx enables you to spend more time on the creative aspects of your design. This helps you get to market faster, and deliver a more robust product to your customers. The
Xilinx 3.1i development systems deliver superior push-button, interactive, and state-of-the-art design methods. The 3.1i release will begin shipping to all registered, in-maintenance customers this spring. To learn more, please visit the Xilinx web site www.xilinx.com.
Logic Analysis with ChipScope ILA
The need for thorough de-bugging capabilities in today’s multi-million gate FPGA designs is critical.
by Shelly Davis, Sr. Product Marketing Manager, Xilinx, [email protected]
T he need for thorough de-bugging capabilities in today’s multi-million gate FPGA designs is critical, and verifying logic externally, by probing package pins or board traces, is becoming increasingly difficult. To solve this problem, Xilinx has created a solution called ChipScope ILA™ that integrates the verification capability into the silicon itself. The new
ChipScope Software, combined with the
Integrated Logic Analysis (ILA) and the
Integrated Control (ICON) cores, allows you to have real-time, full speed access to any node or bus in the chip, with an easy-to-use GUI interface. With these powerful tools, you will spend less time verifying chip functionality and shorten your time-to-market for Virtex, Virtex-E, and
Spartan-II FPGA designs.
Xilinx Partners with Agilent Technologies
Xilinx developed ChipScope ILA with Agilent
Technologies, the industry leader in Logic
Analysis. Output from the ChipScope ILA tool is compatible with the Agilent 16700 series logic analyzers, as well as other industry standards.
Using the Agilent analyzers along with
ChipScope ILA output gives you the possibility of expanding analysis to the board and system level.
The Advantages of Programmability in
Programmable logic provides many unique advantages for on-chip logic analysis compared to fixed logic (ASIC ) solutions. Since ChipScope
ILA cores compiled into a device can be removed after analysis is complete, customers save silicon overhead by not having to permanently lodge test structures on their die. In addition, the cores can be modified and re-placed on the device depending on the signals chosen for analysis. Only a programmable solution allows the user to change the physical silicon. In an
ASIC, if similar cores were included and a subsequent change was required, a new mask set, additional NRE, prototype leadtime and engineering time would be required. For ASIC proto-
Target FPGA with ILA cores
PC with Chipscope
Figure 1 - Integrated Logic Analyzer (ILA) block diagram.
typying, ChipScope ILA provides a highly accurate analysis vehicle that can speed up ASIC development time. For programmable systems, it is truly an innovative method to speed up verification and get to market quickly.
How it Works
ChipScope ILA cores are generated through the
Web, and the ChipScope software can be quickly downloaded to your PC. Once the ILA and ICON cores are integrated into your FPGA design, the
ChipScope software leads you through the process of downloading control information to the
FPGA, modifying trigger and set-up functions, and displaying the captured waveforms. The
Integrated Logic Analyzer core (ILA) provides the triggering and trace capturing functions, and the
Integrated Control core (ICON) communicates to the dedicated FPGA JTAG pins. A MultiLinx cable provides a USB/RS232 interface for communication between the FPGA and the ChipScope software. This is illustrated in Figure 1.
The process works as follows, and is illustrated in Figure 2:
1. The ILA and ICON cores are generated through the Web using a Java application.
2. You insert the cores into your HDL code and connect them to the internal busses and signals that you want to monitor.
Core and ILA Core through the Web
Insert Control and ILA Cores into HDL Code
Connect Busses and Internal Signals
Synthesize Design using
Place and Route the Design with
View and Analyze
Figure 2 - ChipScope ILA flow.
Co-Developed with Agilent
The ChipScope software and ILA core were developed in partnership with
Agilent Technologies, today’s market leaders in logic analysis, and the FBDF format output from ChipScope is com-
3. You synthesize your design, and place and route it using either the Xilinx Foundation
Series or Alliance Series tools.
4. You download the final bitstream to the
FPGA. patible with the Agilent Technology
16700 series analyzer. Visit Agilent at through selecting the correct options.
segment logic and test smaller sections of a large design for greater accuracy.
• Up to 15 independent ILA capture cores per device - Can
5. Your design can then be analyzed through the
• Multiple trigger settings - Records duration and number of events along with matches and ranges for greater accuracy and flexibility.
The ChipScope ILA tools are powerful and accurate, and they provide everything you need to thoroughly verify your logic and make your job easier, including:
• Downloadable from the Xilinx website -
Software and cores easily accessed and downloadable through the Xilinx Xpresso
Cafe e-commerce site.
And More to Come
• User selectable data channels (from 1 to 256 channels) - Accurately captures wide data bus functionality.
• User selectable sample size (from 256 to 4096 samples) - The large sample size increases accuracy and the probability of capturing infrequent events.
• Separate bus trigger (with user selectable width of 1 to 64 bits) - Separate trigger bus reduces the need for sample storage.
• All data and trigger operations are synchronous to your clock (up to 155 MHz) -
Capable of high speed data capture.
In the following months, additional verification tools will be available for use within the
ChipScope suite. The Internal Performance
Analyzer (IPA) core will allow you to log and capture events, and the Integrated Self-Test module (IST) will provide self test capability at the device level. In addition to being available through the Xilinx Xpresso Cafe e-commerce site, the tools will be integrated into the Xilinx
3.1i series software tools.
• Triggers are in-system changeable without affecting your logic - No need to signal step or stop your design for logic analysis.
• Writes waveforms to VCD and FBDF format -
Compatible with Agilent Technologies and other waveform viewers.
• Easy to use graphical interface - Makes learning the software very easy; it guides you
On-chip, full speed, full-featured logic analysis is now a reality. With ChipScope ILA, your designs will be up and running faster than ever and with less effort. For more information see www.xilinx.com or purchase the tools from the Xilinx ecommerce site at www.xilinx.com.
New Products - Software
As Boundary Scan becomes more important as a method for accessing PLDs for programming, robust JTAG tools become essential.
by Frank Toth, Marketing Manager Configurations
Solutions, Xilinx, [email protected]
I n the new Xilinx JTAG Programmer v3.1 software there are significant modifications that improve your productivity and aid in quickly bringing up chains of Boundary Scan (JTAG) devices. Here are some of the new features:
• Chain Integrity Test - This allows you to verify the integrity and robustness of the
Boundary Scan chain to determine its electrical integrity. This is the first step in analyzing why devices in the chain may not be configuring correctly. This test repeatedly reads the
IDCODE register to identify issues related to electrical noise. You can choose the device that will be addressed and then select the number of times to loop and read the IDCODE register. A visual pass-fail indication tells you whether the chain is working properly. By increasing the number of times the device loops through this sequence you can get an idea of the robustness and relative noise immunity of the Boundary Scan chain.
• Readback Capability - This function reads the contents of the FPGA configuration memory and compares it against the bitstream file.
You can use this to verify the correctness of the configuration operations as well as to determine that the appropriate bit stream file was downloaded.
• BSDL Files are Not Required - Now you have the ability to use third-party Boundary
Scan devices with the JTAG programmer software without having their associated BSDL files. All you need to do is indicate the size of the instruction register, which is readily available from the device datasheet. If you have the BSDL file, you can use it as before.
• New Devices Added - Xilinx has paved the way for fully JTAG-compliant PROMs with the introduction of the new XC1800 series. These
IRL-enabling ISP PROMS feature a Boundary
Scan input for downloading along with parallel load (byte-wide) output capability for configuring Spartan devices in Express mode or
Virtex devices in SelectMAP mode. The
XC1800 Proms are all now supported in JTAG
As Boundary Scan grows in importance, push-button, easy-to-use configuration of devices is quickly becoming a necessity. Xilinx has added several new features to the latest edition of JTAG programmer to make Boundary
Scan chain debugging and device programming much easier.
For the latest version of JTAG Programmer, please go to the WebPACK site at: http://www.xilinx.com/sxpresso/webpack.htm.
Software - FPGA Synthesis from
to Implementation in a Xilinx FPGA
The Frontier Design A|RT Designer product efficiently maps a software design into a hardware description language implementation suitable for
by Doug Johnson, Business Development Manager, Frontier
Design Inc., [email protected]
A s FPGA densities soar past the milliongate level, there are enough transistors to put entire systems on a single device.
To help manage these complex designs and reduce simulation time, large systems are typically prototyped, tested and debugged in software that is usually written using the C programming language. C code can execute 10 to 100 times faster than code written in either VHDL or
Verilog HDL, and for complex communications systems C simulation is the often the only alternative.
For hardware implementation, C-language prototypes are usually migrated to an FPGA to enable optimum flexibility and reconfigurability.
The Frontier Design A|RT Designer efficiently and accurately maps a large software design (such as a DSP algorithm written in C code) onto an optimized hardware architecture (FPGA). A|RT
Designer is ideal for creating a processor-like architecture for DSP algorithms such as speech compression, image processing, and for the major blocks in communications systems such as GSM, W-CDMA and IS-136.
Implementing C Code in a Hardware
Algorithms and systems written in C code are sets of sequential operations that have no regard for resources, parallelism, or timing. The sequential nature of software means that clocking and parallelism are not considerations. To get a piece of software into dedicated hardware, specific hardware resources, such as ALUs, multipliers, adders, RAM, ROM, and registers, must be created and the various operations assigned to them. The operations must then be scheduled so that data dependencies are accounted for.
The only way to arrive at the optimum architecture is to look at how the operations are executed on a variety of architectures until the best solution is found.
Until now, there have been no EDA tools that support the exploration of alternative hardware architectures. There was no way to engage in an intermediate architectural exploration between software and silicon. Writing the design in a hardware description language (HDL) requires an explicit architecture because an HDL is a
description of the actual register transfers in a specific architecture. Writing a complex design in a hardware description language is too difficult and time consuming to do more than once.
then generates a register-transfer-level HDL description.
Design Flow Using the A|RT Tools
The EDA software tools from Frontier Design bridge the gap between the complex system as described in C code and it’s final implementation in optimum hardware architecture. Frontier
Design has applied its 15 years of experience in transforming DSP algorithms (written in C code) into working silicon, to create a methodology that is called “Algorithm to Register Transfer,” or
With A|RT Designer, you can interactively add or remove resources, re-assign operations or alter their scheduling, and then analyze the resulting performance characteristics. A|RT
Designer is not constrained by parameters such as clock rate or silicon area, so multiple architectures can be explored very quickly. A|RT
Designer automatically generates a reconfigurable datapath architecture, plus a VLIW (Very
Large Instruction Word) processor and microcode to direct and schedule operations.
Interactive Architectural Synthesis
The focal point of the design flow shown in
Figure 1 is the A|RT Designer tool that takes the
C code and transforms it into synthesizable
VHDL and Verilog HDL that is suitable for driving
FPGA synthesis tools. A|RT Designer automatically synthesizes hardware architectures directly from behavioral-level C-language programs and
A|RT Designer and
Reference Vectors and Stimuli
Although A|RT Designer can be used to automatically synthesize a push-button architecture with minimal intervention by the user, it has been designed with an interactive paradigm from the beginning. A wide variety of reports are available to help you analyze the characteristics of the design. Based on the design constraints, you have a variety of optimization options, plus a variety of directives that you can use to refine the architecture. Design iterations can take as little as a few minutes, so you have the freedom to explore multiple processor architectures very quickly.
Figure 1 - A|RT design flow.
Using A|RT to Target Optimum FPGA
A|RT Designer converts the C behavioral model of an algorithm into an RTL netlist. The resulting hardware architecture consists of a set of data path blocks that are controlled by a microcodebased controller. The datapath blocks are funda-
0 .. n
0 .. n
0 .. n
resource sharing. This means that the tool can reuse the same resource within several clock cycles. For implementation in Xilinx FPGAs, the
LogiCORE blocks can be resources A|RT
Designer uses to build the datapath. The datapath will optimize silicon area and achieve highest performance since they are already optimized for the FPGA architecture.
PC Status and Branch
Automated Controller Generation
ASU=> Application Specific Unit
An ASU is a user defined processing block that can perform an operation such as an FFT butterfly
Figure 2 - A|RT designer processor architecture.
The benefits to the communications system engineer and DSP algorithm designer and hardware engineer are enormous because the
LogiCORE and Core Generator blocks are now usable by both the system engineer and the hardware engineer. The system engineer can now incorporate pre-defined hardware functionality directly in C code. The hardware engineer gets a VHDL or Verilog HDL model from A|RT
Designer that directly instantiates area and performance-optimized LogiCORE and Core
Generator DSP building blocks. mental DSP building blocks such as multipliers,
ALUs, register files, RAM, and ROM. Figure 2 is a simplified illustration of the processor architecture used by A|RT Designer.
One of the most powerful capabilities of the
A|RT methodology is the ability to target predefined, optimized components in the C code, such as those created by the Xilinx Core
Generator and those in the Xilinx LogiCORE library. By exploiting these predefined building blocks, a highly optimum, silicon-efficient implementation of the architecture is possible. For each of the LogiCORE blocks, it suffices to create a simple model file for the resource, containing information on the I/O pins, the cycle-level timing, the instructions available on the resource, and the C functions that can be mapped to those instructions. In addition, A|RT Designer performs
Also, a DSP algorithm typically requires extensive memory such as register files or block
RAM and the tools can be directed to target the block RAM capabilities of the Spartan and Virtex devices extending the efficiency of the target
A|RT Designer is supported for PC Windows/NT 4.0, for HP-UX 10.20, and for Sun/Solaris 2.7. For more information, please see the Frontier
Design website at www.frontierd.com.
New Products - Cores
New DES and Triple DES cores meet the requirements for highperformance systems as well as smart cards, cable modems, and
Bluetooth wireless systems.
by Amit Dhir, Sr. Engineer, Strategic Applications, [email protected]
A s more companies conduct business over the Internet, transaction security is a major concern. Therefore, to help pave the way for secure e-commerce, Xilinx and
Xentec, Inc. recently announced the availability of two new core products for Virtex, Virtex-E, and Spartan-II FPGAs. The Xentec encryption cores, along with the inherent flexibility of Xilinx
FPGAs, makes it easier for you to create secure data transmission systems.
Xentec’s X_DES core is certified by NIST to conform to the FIPS 46-3 and ANSI X9.52 specifications. (See http://csrc.nist.gov/cryptval/ des/desval.htm for more information). Both the
X_DES and X_3DES cores contain encryption and decryption functions selectable by internal control. In secure data communications, the same core is used at both transmit and receive ends.
The National Institute of Standards and
Technology (NIST) currently recommends the use of Data Encryption Standard (DES) and triple DES (3DES) cryptographic algorithms for data security. The DES function encrypts 64-bit data using a 64-bit secret key. Authorized users can only decrypt data with the same key.
Triple DES is a cascaded chain of three single DES functions to further enhance the security.
The X_DES core supports all four standard
DES modes: Electronic Code Book (ECB), Cipher
Block Chaining (CBC), Cipher Feedback (CFB), and Output Feedback (OFB). The X_3DES triple
DES core supports the most popular ECB mode while customization is available for other modes. Both cores process a new encryption or decryption every 16 clock cycles.
“Systems which require flexibility and highperformance integration will benefit from the
Virtex series version, while cost-sensitive, highvolume consumer applications will benefit from the Spartan-II versions” said Vincenzo Liguori, design manager of the multimedia group at
Use the Virtex Series for
Virtex series FPGAs are well suited for high-performance applications due to their high clock speeds, large gate densities, and system-level fea-
Xentec, Inc. provides comprehensive ASIC and FPGA design services, integration expertise, and technology for the product development requirements of the world’s leading electronics companies. Founded in 1995,
Xentec’s mission is to be the leading provider of analog/digital integrated circuit design services and Intellectual Property used in
System-on-Chip (SoC) based integrated circuits for all facets of the electronics industry.
The company is headquartered in Oakville,
Ontario and is privately funded. More information about the company, its products, and services may be obtained from the World
Wide Web at www.xentec-inc.com. For inquiries regarding Xentec cores and services please contact [email protected]
tures. The X_DES core runs at an effective bit rate of about 500 Mbits/sec in
Xilinx Virtex-E devices. immediately available for use in Xilinx Virtex series and
Spartan-II FPGAs. The netlist versions of the X_DES and
X_3DES triple DES cores list at $7,500 and $10,000 respectively. All Xentec products can be purchased directly from Xentec; the data sheets can be downloaded from the
Xilinx IP Center (www.xilinx.com/ipcenter), a comprehensive resource for system-level intellectual property and services.
You can easily integrate the small DES and triple-DES cores with other Xilinx IP solutions to build large complex systems in Virtex series
FPGAs. These include Internet, intranet, and extranet networks facilitating secure businessto-business transactions, satellite digital cinema, secure satellite broadcast, secure video surveillance, and space imaging systems.
Use the Spartan-II Series for Consumer
Data Encryption Cores
The DES and triple-DES cores are very compact.
The cores targeted for the Spartan-II FPGA family offer a low-cost solution that you can use in combination with other Xilinx IP solutions to develop consumer appliances that include ecommerce security enabled PCs and cable modems, set-top boxes, wireless LAN and
Bluetooth wireless systems, prepaid smart cards, and personal banking systems.
Whether you are creating low-cost, high-volume consumer equipment or high-cost, high-performance systems, its very easy to incorporate a high level of data security into your designs. The combination of Xentec DES cores and Xilinx
FPGAs gives you a very flexible solution that will get your secure products to market as soon as possible.
Pricing and Availability
The cores are sold and supported by Xilinx
AllianceCORE™ partner, Xentec, Inc. of Ontario,
Canada. The X_DES and X_3DES cores are
New Products - Cores
These new cores target JPEG, MPEG, DSP, and image processing applications.
by Antolin Agatep, [email protected], Systems Architect, Embedded Systems
X ilinx and Xentec recently announced new
AllianceCORE products for Image processing. These cores include a discrete/inverse discrete cosine transform (DCT and IDCT) and a JPEG codec. The DCT/IDCT core supports Virtex, Virtex-E, and Spartan-II devices, and is a critical building block in Dolby AC2 and
AC3, JPEG, and MPEG compression systems. The
JPEG codec supports Virtex and Virtex-E devices, and provides a fully integrated encoder-decoder pair used in image compression and decompression applications. blocks of JPEG, MPEG, and ITU-T H261 standards-based codecs that are used in many image processing applications.
A DCT function is a key element in a compression system that splits still-image pixels into smaller data blocks. It calculates a value to represent each block that can be used to reduce the storage space required for the overall image. The
IDCT function operates in reverse and reconstructs the image from compressed data.
Target applications for the new cores include image storage, data and video compression, medical imaging, and industrial systems. Xentec developed these very compact image compression cores by using the versatile DSP features in the Spartan-II and Virtex architectures.
Xentec’s DCT/IDCT core performs both DCT and IDCT functions and is ideally suited for systems that require both image compression and decompression. It supports both DCT and IDCT on 8x8 image pixel data. DCT or IDCT mode is selected by means of a control signal. The core fits in a single XC2S100 Spartan-II FPGA that costs just $10 in high volume.
DCT and IDCT Solution in a $10 FPGA
The DCT/IDCT core enables high-speed hardware implementation of the forward and inverse discrete cosine transform functions. DCT and
IDCT (Inverse DCT) functions are the building
“Digital imaging applications typically need the power of 32-bit processors for implementing the computationally intensive DCT/IDCT functions in software,” said Robert Bielby, Xilinx
Director of Strategic Applications. “By moving the DCT/IDCT function into a cost-effective
Spartan-II device, designers can now use cheaper eight-bit microcontrollers and cut down the overall system costs.” offers unlimited programmability and can be upgraded in the field.
Integrated JPEG Codec Solution
JPEG is an image compression technique for reducing image file sizes without a noticeable difference in image quality. The Xentec X_JPEG codec core conforms to the ISO/IEC 10918-1
JPEG baseline specification and performs both compression and decompression functions. The full-featured core includes support for stalling and the ability to handle four-color components.
It is extremely flexible, including four programmable quantization and four programmable
Huffman tables that allow you to specify quantization, Huffman table, and the quantity of pixel blocks assigned for each color component.
“The memory hierarchy in the Xilinx Virtex-E
FPGAs enabled us to develop an extremely small
JPEG core,” said Xerxes Wania, President and
CEO of Xentec. “We used distributed
SelectRAM™ to build ROM-based Huffman tables that are distributed across the design and embedded block SelectRAM for the large
DCT/IDCT, quantization, and zig-zag coding tables. The end result is a flexible, optimized codec solution that includes both logic and memory in a single medium-density FPGA.”
The Spartan-II family delivers 100,000 system gates for under $10 at speeds of 200 MHz and beyond, and provides unmatched design flexibility. These low-power 2.5-volt devices feature I/Os that operate at 3.3 volts with full 5-volt tolerance. Spartan-II devices also feature multiple Delay
Locked Loops (DLLs), on-chip RAM
(block and distributed), and the versatility of SelectI/O™ technology supporting over 16 high-performance interface standards, in a device that
The Virtex-E family has made significant improvements in the areas of capacity and performance over the original Virtex series. The family features devices with 3.2 million gates,
832 Kbits of True Dual-Port internal block RAM, eight digital Delay Locked Loops (DLLs) capable of over 300 MHz clock frequency for system timing, and three new differential signal standards.
The Virtex-E FPGAs are the first programmable logic devices delivered on 0.18-micron process technology, which Xilinx jointly produced with
Taiwan’s UMC Group. The improved process directly contributes to the 30 percent performance gain for all internal functionality. Also, the Virtex-E family represents the industry’s first programmable logic architecture with 210 million transistors on a single device.
Availability and Pricing
The X_JPEG and X_DCT_IDCT cores are immediately available from Xentec, Inc. The DCT and
IDCT core is available for use in Spartan-II,
Virtex-E, and Virtex FPGAs, and lists at $10,000 for the netlist version. The JPEG codec is available for use in Virtex-E and Virtex FPGAs, and lists at $30,000 for the netlist version. All Xentec products can be purchased directly from Xentec.
The datasheets can be downloaded from the
Xilinx IP Center (www.xilinx.com/ipcenter), a comprehensive resource for system-level intellectual property and services.
Xilinx FPGAs, along with the Xentec cores, provide a complete solution for creating image processing applications. This combination helps you quickly get your products to market, saving you time and money.
Success Story - Cores/QPRO
Using an 8051
A Swedish company develops unique military products using Xilinx
VirtexXCV300 FPGAs and a high speed 8051 core.
by Oscar Blanquez, Project Manager, Dolphin Integration, S.A., [email protected],
Lars Albihn, Sr. Design Engineer, CelsiusTech Electronics, AB, [email protected],
Anil Telikepalli, Sr. Technical Marketing Engineer, Xilinx, Inc., [email protected]
C elsiusTech Electronics AB specializes in advanced electronic military systems. In
May 1999, they evaluated the feasibility of using FPGAs and IP modules for their new designs. Five microcontrollers were evaluated, and the Dolphin Flip8051-PR AllianceCORE product along with Xilinx Virtex QPRO FPGAs were selected as the perfect solution.
(see Figure 1). The system receives, detects, identifies, and warns the pilot of hostile radar signals without warning the enemy. The frontend unit of the RWR system needs a compact and reliable IP module control processor, such as the 8051 family microcontroller. Since this frontend unit is mounted external to the supersonic aircraft, it is subjected to severe mechanical, thermal, and electromagnetic conditions.
Dolphin’s Flip8051-PR is available for use in the Xilinx Virtex, Virtex-E, Virtex QPRO, and
Spartan-II devices. The Virtex QPRO family is fully qualified for military use and offers a wide range of devices up to 2-million gate densities.
The Flip8051-PR 8051 microcontroller core is known as the fastest 8051 Virtual Component for
ASIC applications, running an average of eight times faster than the legacy i8051, and it gives you enormous flexibility, something a standard device can’t offer. This combination of speed, flexibility, density, ease of use, and military qualification was an easy choice.
The RWR system includes the RWRW frontend unit, Central RWR Computer, and Ground
Maintenance Computer (see Figure 1). The frontend unit is composed of:
• FLIP8051-PR IP module with internal ROM and RAM integrated into the FPGAs using block RAM and distributed RAM features.
• Microwave Receiver Subsystem, DSP Function block.
• Two serial links: a standard 8051 UART and an in-house High Speed Serial
Developing a Radar Warning Receiver
Easy Integration of the Core
One of the systems currently being developed by
CelsiusTech is a Radar Warning Receiver (RWR) system, used in a supersonic combat aircraft
The Flip8051-PR core was easy to implement in a Virtex XCV300 FPGA and it was successfully integrated with all the sub-modules of the RWR
Information from other sensors
RWR Front-end Unit
High-speed serial lnks
Using the Bus Monitor for
system within a short time. The core provided key advantages such as:
Figure 1 - Radar Warning Receiver (RWR) system.
• Code compatibility with legacy i8051.
Before implementing the applications in the FPGAs, the systems were fully simulated and verified using a Bus Monitor function provided by Dolphin. The Bus Monitor interacts with a simulation tool such as Model Technology’s V-System and generates text messages of disassembled instructions so errors can be observed and easily fixed (see Figures 3 and 4).
Disassembled Instructions and micro-controller status display in simulator
• Easy access to Special Function Register (SFR) space.
• Demultiplexed address/data bus for memory interface.
• Choice of bidirectional/unidirectional I/O.
• Bus Monitor function for simulation.
Bus Monitor Port
The design team at CelsiusTech found the core’s SFR Bus Interface especially helpful because it allowed them to easily add peripherals to the core (see Figure 2). The SFR bus gives a lot of flexibility, especially when designing systems that include many on-chip peripherals.
Figure 3 - Typical use of the Bus Monitor functionality.
Data from peripherals
Data to Peripherals
Figure 2 - SFR Bus Interface.
CelsiusTech has also used the FLIP8051-PR in several other applications, incorporating highly application-specific peripheral functions.
The Flip8051-PR core has a dedicated port used by the Bus Monitor behavioral model for tracing code execution and disassembly in real time. In addition, the dedicated port monitors the microcontroller interrupt status.
Once simulations were verified, the design was downloaded into the Xilinx Virtex V300
FPGA and integrated into the evaluation board
(see Figure 5). The core worked successfully on the first attempt and the CelsiusTech team was able to incorporate additional peripherals on the same evaluation board.
“The need for support from
Dolphin has been less than originally anticipated, mainly because the FLIP8051-PR has proved to be a mature design of high quality,”
Figure 4 - Screen dump of Bus Monitor
Model Technology’s V-System.
The simulator facilitates simulation of both standard 8051 peripherals and user specific peripherals.
Figure 5 - CelsiusTech Evaluation Board.
“The need for support from Dolphin has been less than originally anticipated, mainly because the FLIP8051-PR has proved to be a mature design of high quality,” said Lars Albihn, senior design engineer at CelsiusTech. By using a proven AllianceCORE product from Dolphin in a
Xilinx FPGA, CelsiusTech was able to save both time and money while benefiting from the technical expertise of Dolphin.
Dolphin has partnered with Raisonance
(www.raisonance.fr) to provide a full software tool suite, the Flip8051-Rkit. The tool suite consists of a C compiler, an assembler, a linker, a real time OS, a ROM monitor, and a simulator.
Based on the successful evaluation and deployment of the Flip8051-PR core in their systems,
CelsiusTech concluded that using IP modules is the best method for their system design on
FPGAs. Thanks to features such as high-speed execution, SFR Bus Interface, and Bus Monitor, the FLIP8051-PR in a Xilinx FPGA has proved to be a reliable, flexible and easy to use Virtual
Component. It is an ideal solution that adds the expertise from Dolphin to the flexibility and system level features of Xilinx FPGAs thus, reducing the overall time-to-market.
For more information and a datasheet of the Flip8051-PR core, see the
Xilinx IP Center, a comprehensive resource for system level intellectual property and services: www.xilinx.com/ipcenter
Dolphin Integration, is a Xilinx AllianceCORE partner in Europe with expertise in digital, analog, and mixed signal circuits. They provide IP solutions for processors, telecom, and multimedia systems.
Dolphin Integration, 39, Avenue du Granier, BP 65 Zirst, F-38242 MEY-
LAN Cedex, FRANCE, www.dolphin.fr
CelsiusTech Electronics AB, Nettovägen. 6, JAKOBSBERG, SE-175 88
JÄRFÄLLA, SWEDEN, www.celsiustech.se/electronics
New Products - High Speed DRL
5 Gbytes/sec digital I/O DIME Module now supports DRL technologies.
by Derek McAulay, Design Engineer, Nallatech Ltd, [email protected]
N allatech Ltd. recently announced another new addition to its DIME
Standard family, a revolutionary new module named “Ballyderl” that provides an intelligent, dynamically reconfigurable platform with high speed off-board communication. Based upon the DIME plug-and-play capability, this module harnesses the Virtex system-level features and enables you to quickly build multiple
Using the latest data transmission technologies, you can attain clocking rates over 311 MHz, collectively allowing a data transmission and reception rate of 2.4 Gbytes per second. This complements the 2.5 Gbytes/sec capability that is currently available on other DIME modules.
LVDS or Single Ended
LVDS or Single Ended
Virtex / VirtexE FPGA
XCV600 up to
16M x 32 (64
16M x 32 (64
Figure 1 - Ballyderl block diagram.
low power consumption and low susceptibility to common mode noise and EMI effects.
• Single ended. This option allows you to interface with standard LVTTL (TTL for Virtex build) enabling this module to communicate with other systems that do not support LVDS.
If you need 5-V tolerance or support for the current Xilinx DRL technologies such as JBits and XVPI, you can also specify a standard Virtex device, instead of a Virtex-E device.
DIME: DSP and Image Processing Module for Enhanced FPGAs.
The Ballyderl module incorporates a wide range of Virtex and Virtex-E FPGAs, from the XCV600 to the XCV2000E, giving maximum flexibility that is further enhanced by a variety of digital I/O transmission options and two independent 64
Mbytes banks of SDRAM, as illustrated in
The easy to use Ballyderl DIME module allows you to configure designs in minutes, and provides high performance reconfigurable computing and data processing with unrivalled efficiency.
The module features two build options for transmitting and receiving the 64 digital I/O signals:
• Low Voltage Differential Signaling (LVDS).
This option allows you to pass your data over long distances at high rates while maintaining
For more information on Ballyderl, or other DIME products, contact Nallatech at: www.nallatech.com.
Other DIME Modules include:
Ballyblue: Dual 2 million VirtexE DIME Module.
Ballyvision: Analog video I/O DIME Module.
Ballysharc: VirtexE, Dual Hammerhead SHARC DSP DIME Module.
Ballyriff: 4 Channel 105MSPs 12bit ADCs, VirtexE DIME Module.
New Products - Support
64/66 Design Kit
Includes a prototyping board, driver development tools, and reference designs.
by Per Holmberg, Product Marketing Manager IP Solutions,
Xilinx, [email protected]
X ilinx recently announced The Real 64/66
PCI™ Design Kit, which includes access to the proven Xilinx LogiCORE PCI 64/66 and PCI 32/33 cores, a 3.3-V/5-V universal 64bit/66-MHz PCI prototyping board, Windows driver development tools, and several PCI
Reference Designs. The kit provides the only complete solution for developing 64-bit, 66-MHz fully compliant PCI add-in boards for the PC, server, and workstation markets.
When Xilinx introduced The Real
64/66 PCI in March, last year, it was the industry’s first general-purpose
64-bit, 66-MHz PCI solution. This new
PCI standard has until now mainly been adopted by companies developing embedded applications. However, with new high-performance PC motherboards entering the market with support for 64-bit/66-MHz PCI, add-in board manufacturers for PCs, Servers, and Workstations started developing products using this new PCI standard.
With this design kit, Xilinx extends its solution to include all tools required to support this emerging market of high-performance PCs with
64/66 PCI capabilities. Not only will the Real
64/66 PCI Design Kit accelerate your time-tomarket, but it also addresses the emerging need for concurrent support of 5-V PC motherboards and the next generation, high-performance 3.3-V
To provide a complete solution and industry leading expertise, Xilinx has partnered with
Compuware, the leading provider of software
driver development and debugging tools, and
Nallatech, a member of the Xilinx third-party
XPERTS consulting program, providing PCI design services and the prototyping board.
The NuMega SoftICE Driver Suite
With The Real 64/66 PCI design kit, you receive a production version of Compuware’s NuMega
SoftICE Driver Suite. These tools accelerate the development and debugging of Windows device drivers, offering a full-featured solution for basic driver development and debugging in Windows
2000, Windows NT, and Windows 95/98 platforms. A standard license for the Compuware suite is included.
“NuMega SoftICE Driver Suite addresses a critical need for driver developers seeking highquality development tools to increase their productivity,” said Joe Wurm, vice president and lab director for the NuMega product line. “This release of the SoftICE Driver Suite enhances the debugging and testing capabilities of the suite to fully encompass Windows 2000 driver develop-
“Not only will the Real 64/66
PCI Design Kit accelerate your time-to-market, but it also addresses the emerging need for concurrent support of 5-V PC motherboards and the next generation, high-performance 3.3-V
ment. The SoftICE Driver Suite is a key element of Compuware’s comprehensive strategy for helping corporations deploy reliable Windows device drivers.”
The PCI64 prototyping board
The Nallatech 64/66 PCI prototyping board allows you to quickly evaluate the performance of the Xilinx core in your system. In addition, the
board demonstrates how to build a universal PCI board that automatically changes the I/O characteristics as needed to support both 3.3V and 5V PCI.
Additionally, this board shows how to meet the stringent PCI specification requirements for sub-100ms PCI reset timing with
Nallatech is a leading electronic systems design company that specializes in the development of highly complex DSP and image processing systems for its customers. Nallatech has also developed a modular approach to the construction of highly complex systems using their innovative DIME (DSP and Image processing Module for Enhanced FPGAs)
Products. Founded in 1993 and headquartered in Scotland,
Nallatech is bringing the advantages of the FPGA directly to system engineers, allowing them to construct complex systems with a minimum of risk. DIME and Xilinx have finally brought to hardware design the flexibility that software has enjoyed for many years. For more information, visit the
Nallatech website at http://www.nallatech.com.
An SDRAM interface to a Small Outline
DIMM (dual inline memory module), using the integrated digital Delay Locked Loops of the
Xilinx Virtex FPGAs, demonstrates how to build a single-chip solution. With Virtex FPGAs, boardlevel DLLs are not required because the FPGA can handle clock management, both on and off the chip.
compliance is verified through hardware testing, device characterization, and regression testing using the Xilinx internal test bench that simulates more than six million unique combinations of PCI transactions.
With its LogiCORE offerings, Xilinx has enabled thousands of
32-bit and 64-bit PCI designs. This translates to more than twice the revenue generated by the leading chip-set competitor. This is a significant milestone that underscores the inherent benefits of standard FPGAs on the most advanced processes. More important, The Real 64/66 PCI solution allows you to integrate very high-performance, high-density 66-MHz PCI systems tailored to your specific needs.
The Xilinx Real
64/66 PCI Core
The Real 64/66 PCI core from Xilinx enables you to design fully compliant PCI bus interface systems, using Xilinx Smart-IP technology to guarantee the critical minimum, maximum, and hold timing required for a true zero waitstate burst operation at 66 MHz. PCI v2.2
With trailing 12-month revenues of more than $2 billion,
Compuware is a world leader in the practical implementation of enterprise and e-commerce solutions. Compuware productivity solutions help 14,000 of the world’s largest corporations more efficiently maintain and enhance their most critical business applications. Providing immediate and measurable return on information technology investments,
Compuware products and services improve quality, lower costs, and increase the speed at which systems can be developed, implemented and supported. Compuware employs more than 15,000 information technology professionals worldwide. For more information about Compuware, please contact the corporate offices at 800-521-9353. You may also visit Compuware on the World Wide Web at www.compuware.com. For information on Compuware’s SoftICE Driver
Suite please visit www.compuware.com/drivercentral.
It’s now easier than ever to design fully compliant PCI-compatible systems. The PCI design kit with the
Nallatech prototyping board and Compuware software is available now for $17,995 from
Xilinx distributors. For more information see www.xilinx.com/pci.
New Products - Support
The Configurator is a micro-controller with Flash memory, allowing configuration of up to eight FPGAs in parallel.
by Greg Clifford, Marketing, Configurator, Inc., [email protected]
T he Configurator from Configurator, Inc., shown in Figure 1, provides a large FLASH memory to store configurations for all of the Xilinx FPGAs. It also includes an intelligent link to your PC along with control software, so you get enough storage space for multiple devices and multiple configurations with fast and easy file management.
The Configurator allocates each output of the
FLASH memory to a Xilinx FPGA, permitting up to eight FPGAs to be configured at once. If fewer than eight FPGAs are to be configured, the
Configurator will automatically use more than one data bit for storage for each FPGA, thereby increasing the bitstream file storage space available.
The Configurator supports all XC3000,
XC4000, XC5000, Spartan, and Virtex devices with 8-, 16-, and 32-Megabit models available.
The Configurator is programmable in-system using a Windows application program communicating over a 115k Baud serial port connection.
It’s available now in a 32-pin DIP package for easy integration into circuit card designs.
The selection of a Configurator model’s flash memory size is based on the number of FPGAs to be configured and the size of each FPGAs bitstream file. For example, a Configurator-8M (8 megabits) can configure up to eight FPGAs with each one having a maximum bitstream file length of less than one megabit. However the same Configurator-8M will automatically interleave bits of data so that if only four FPGAs need to be configured on the card, the maximum bitstream file length is increased to two megabits each.
Figure 1 - Configurator 32-pin DIP module.
Table 1 illustrates a subset of the storage capabilities provided by a single Configurator. To use this table:
1. In the left hand column select the largest
FPGA you will be configuring.
2. Go right across the top selecting the number of FPGAs and the number of configuration planes.
3. Where row and column intersect is the
Configurator model you will need. If you have an odd number of FPGAs to configure or an odd number of configuration planes round up to the next power of two.
For example, if you have four FPGAs to configure, the largest of which is a Virtex XCV50, and you want three different configuration planes, go down the left side until you get to the
XCV50 row then across to the “4 FPGAs/ 4 planes” column indicating that a Configurator-
8M would satisfy your requirements.
The Configurator adds on-board configuration control to FPGA designs by providing the capability to store custom configuration information in the flash memory. The date, path, and filename are stored automatically and user information such as design version, firmware numbers, and so on, may be stored with each bitstream file and are user configurable. This configuration information is especially helpful during FPGA development to identify revisions of bitstream files.
Figure 2 - Windows configurator user interface.
Figure 2 runs on Windows 95/98/NT systems and provides an easy interface to the
Configurator. To download new bitstream files to the Configurator, connect to a serial port, select the .bit files, then select “Program.” The GUI will download the configuration files at 115K bits per second, saving the data to FLASH memory. For example, if an additional signal probe is needed during debugging, the power may be turned off, the probe attached, and when the power is turned back on the Configurator module will reconfigure the FPGAs without the need for another download.
The Configurator also minimizes field upgrade costs by providing in-system re-programmability, a simple RS-232 serial port interface, and configuration information to help field service personnel determine and update the required bitstream file changes.
Internet Reconfigurable Logic
The Windows GUI communicates with the
Configurator module using ASCII character protocols providing an easy method of performing remote bitstream file updates. The formatted bitstream files are also text files that are easily distributed. The communication link could be an
Internet connection permitting hardware updates from anywhere in the world.
Windows User Interface
The Configurator Windows GUI as shown in
Re-configurable Computing Applications
The Configurator has the ability to store configuration files in up to eight different planes. Each plane can have one configuration file for up to eight FPGAs. The planes are selectable from the host card’s hardware to support re-configurable logic applications. This feature permits multiple
Figure 3 - Configurator example schematic.
Table 1 - Configurator storage capability.
functions to be implemented for the design utilizing the same hardware. schematic in Figure 3 shows a Configurator connected to Virtex, XC4000, and XC3000 FPGAs.
For example, one plane could have a built-in self-test function, while the next plane might have a data process function, and the third plane being a data compression function. Having multiple planes simplifies the design since all of the functions do not need to be accommodated in a single FPGA design and a smaller and faster
FPGA size is realized. The Configurator has three dedicated input pins to allow for the configuration plane selection, the planes are selected by setting the RECONFIG[2:0] inputs to the desired plane followed by resetting the device.
You can buy a self-contained module to install on each board or the Configurator can be integrated into your board design by purchasing just the pre-programmed Configurator microcontroller. The Configurator Starter Kit is $459, which includes a Configurator Module, a serial interface cable, sample schematics and the
Configurator Windows GUI. Individual 32-pin DIP modules are available in 8-, 16- and 32-M bit models and pre-programmed micro-controllers are available for higher volume applications.
Configurator Integration Options
Adding the Configurator to your designs is easy, and even existing designs can use a Configurator if the design has an Xchecker cable connection.
The Configurator typically replaces both serial
PROMs and the Xchecker cable connections on a board for up to eight FPGAs. The sample
The Configurator makes it very easy to program
Xilinx FPGAs and it supports the Xilinx Internet
Reconfigurable Logic capability allowing you to remotely reconfigure Xilinx FPGAs, over the
Internet, anywhere in the world.
For further information, visit the Configurator, Inc. website at www.fpgaConfigurator.com.
Using the Virtex
The Virtex Look-up Tables have some interesting capabilities that allow you to create very fast and efficient designs.
by Marc Defossez, FAE, Xilinx BeNeLux,
X ilinx FPGAs have always had combinations of Look-up Tables (LUTs) and flipflops, combined into Configurable Logic
Blocks (CLBs). With the introduction of the
XC4000 family, the Xilinx LUTs also have RAM and ROM capability. Now, with introduction of the new Virtex architecture, LUTs can also be used as shift registers.
Figure 1 - LUTs used as RAM, ROM, or SSR.
The Virtex LUT
The Virtex LUTs have four inputs and one output, and can be used as RAM, ROM, or Serial
Shift Registers (SSRs):
• Used as a look-up table the LUT can contain (up to) any four-input function.
• Used as RAM or ROM the LUT can have a
• Used as an SSR the LUT can be used as a
16-bit shift register.
Creating Counters in LUTs
A counter, whether it’s binary, Johnson ring, or
LFSR, is a sequence of repeating patterns. If you program that sequence into a set of Look-up
Tables, then you can easily create fast, simple, and large counters. Here are several examples.
A 4-Bit Binary Counter
A 4-bit binary counter has 16 possible states which can be stored in a Look-up Table. By cycling through the addresses you can generate this binary counter pattern as shown in Figure 2.
When we initialize an SRL16 with these values and cycle the shift register, the output will behave as a counter. However, the shift register will be empty in 16 clock cycles so you need to connect the input to the output to make the counter repeat the cycle. It’s also possible to give
Binary Counting Sequence
When we turn this counting sequence, then we get:
A LUT, RAM or SRL is 1 bit by 16.
Here we have i4 bits by 16.
So we have 4 LUTs that can be INITialized with hard coded values:
LUT 3 gets 00FF = Q3
LUT 2 gets 0F0F = Q2
LUT 1 gets 3333 = Q1
LUT 0 gets 5555 = Q0
Figure 2 - 4-bit binary count sequence.
this counter a terminal count. To do this, use the carry chain as a wide AND gate and combine all the outputs to create the desired terminal count, as shown in Figure 3.
Using four LUTs, configured as an SRL16 shift register, can make a small 4-bit counter with a full and repeatable sequence.
You can also make a counter or sequencer that generates any count sequence you like, as shown in Figure 4. By initializing the SRL16 with a sequence, so that the counter starts counting at a specific value and stops at another value, then you only need to set the address of the
SRL16 to the correct number of stages.
Figure 3 - Creating a terminal count.
Figure 4 - A 6-bit counter.
Figure 5 - A loadable counter.
A 5-Bit Binary Counter
To create a 5-bit counter, initialize the extra LUT
(SRL16) to “0” and feed it’s output back to it’s input via an inverter. Then for the first 16 cycles the LUT/SRL will give a zero at the output, while loading a “1” into the SRL/LUT. For the next 16 cycles the counter will produce a “1” as the high order bit.
A 6-Bit Binary Counter
To create a 6-bit counter, first create a 5-bit counter as previously described. Then, to create the sixth bit, initialize the LUT to all zeros.
Connect the output of the SRL to it’s input via an inverter and then connect the enable of the SRL to the output of the previous bit, as shown in
The sequence for both upper bits will then be as follows:
1. Both upper bits will be “0.”
2. After 16 clock cycles, bit 5 will become a “1” and bit 6 stays “0.”
Figure 6 - Example of an LFSR counter.
Figure 7 - Simulated counter waveforms.
Figure 8 - Simulated counter waveforms, zoomed in.
3. The LUT of bit 6 will be enabled for the next
16 clock cycles.
4. The LUT of bit 6 will output 16 zeros and then load a “1.” Bit 5 will load “0” while the output is “1.”
5. After 16 clock cycles, bit 5 will become “0” and bit 6 will be a “1.”
The simulated counter is shown in Figure 7 and Figure 8.
A Loadable Counter
Figure 5 shows how to create a loadable counter. A load operation will take 16 clock cycles, and you can load the counter while it is counting.
A Large LFSR Counter
An LFSR counter is a shift register with it’s input fed back (XORed) from the bits of the different stuck states that can appear in the sequence. In the Virtex architecture we have the SRL16 elements that represent 16 flip-flops, thus large
LFSR counters with only certain outputs of interest can be made efficiently.
For more information see Application note (Xapp052) on “Efficient
Shift Registers, LFSR counters, or see Xcell 35 article on “Pseudo
Random Noise Generators.”
Application - Software
You can easily view internal signals by routing them to device pins.
by Davin Lim, Software Technical Marketing Engineer,
Xilinx, [email protected]
T he PROBE capability within the
FPGA_Editor application (included in all
Xilinx Foundation Series and Xilinx
Alliance Series development systems), allows you to quickly identify and route any internal signal to an unused I/O pin. Once routed to the pin, you can then view that signal’s real-time activity using normal lab test equipment such as logic/state analyzers and oscilloscopes. PROBE is very easy to use and it makes your lab debug cycles as quick and efficient as possible.
Your HDL design effort only needs to focus on the device’s primary intended function because all of PROBE’s power can be accessed automatically during the lab-debug phase of your project design cycle—there is no need to guess before lab debugging which signals you wish to analyze. PROBE easily guides you through the debug phase, so you can pick which signals need to be observed at any time during the process, and changes to the signal list can be made instantaneously.
PROBE gives you a simple, natural way to observe any internal signals in your design and analyze their switching behavior with essentially no impact on design performance. The signals are viewed in real-time, so you are sure to have the most accurate picture possible on how your design is functioning in it’s actual operating form and environment.
The PROBE function can be used either in manual or automatic mode.
As shown in Figure 1, using PROBE is as simple as loading the design into FPGA_Editor and clicking on the “PROBES” button on the User
cially in cases where ball-grid array
(BGA) packages are involved, you will connect test points on your PC board to specific FPGA device pins. In such cases, you’ll know exactly which unused I/O pins are most convenient to use for
PROBE connections, and the manual mode is the best choice.
Figure 1 - How to access PROBE.
When reserving the I/O pins you wish to use as test points, you can increase the effectiveness of PROBE by selecting at least one I/O pin on each side of the FPGA (as viewed in the
Floorplanner or FPGA Editor). This way, no matter where on the FPGA the internal signal to be probed exists, there will be a test I/O pin located nearby.
Once the PROBES dialog box appears, all you need to do is pick the signal to be analyzed and assign it to an unused I/O pin, as shown below in Figure 2.
In automatic mode, PROBE automatically chooses an unused I/O pin that will incur the least amount of routing delay when connected to the signal you want to monitor. To use the auto-
Choose signal to
matic mode, simply select “Automatic” under the
“Method” selection box. Once the signal, pin list, and pin selection method have been chosen, pressing the OK button will complete the PROBE operation.
Assign to unused
Displaying PROBE Results
Figure 2 - Assigning signals to probes.
This is the most commonly used mode because most of the time you will know which
I/O pins are best to use for connection to external logic analysis equipment. Usually, and espe-
The Probe list window lists all signals that have been assigned to PROBE I/O pins. This list can be sorted by any of the headings simply by clicking on the desired column’s heading. It will often be important for you to know how much routing delay is incurred to route the signal to the I/O pin. This delay amount is listed in the PROBE list
The amount of route delay on the actual probed signals is often not critical to the debug task at hand. As long as you
Figure 3 - Routing delay of probed signal.
know what the route delay is, you can account for that amount in the analysis process, and therefore it may not be necessary to always optimize the route delay to under the “Slack” heading. In the example shown in Figure 3, there is 3.6ns of routing delay to route the signal “mult4/ADDER_01” to the external I/O pin.
achieve minimum delay. If this is the case for the signals you are probing in your design, you can significantly quicken the PROBE execution process by simply ensuring that the .pcf file is not
Controlling Probe Execution Time and
read in when the design is originally loaded into
FPGA_Editor. This is accomplished by making
If your design has timing constraints which specify timing requirements for items such as clock periods and I/O timing, the default routing algosure the Physical Constraints File field is left blank, when completing the File->Open dialog box as shown in Figure 4.
Speed up PROBE routing time by loading design without Physical
Make sure this field is blank.
To quickly and effectively get through the lab-debug phase of your FPGA design, you will need tools that are fast, easy to use, and work with a minimum of fuss. The
PROBE capability within the FPGA_Editor application directly addresses this need.
PROBE provides you with direct access to the internal signals of your design, with the absolute minimum impact on the design’s performance and resource utilization.
Figure 4 - Omit the PCF file to speed routing times.
rithms used by PROBE will account for those timing specifications and work to ensure that all timing constraints are still met after the probe signal has been inserted. FPGA_Editor accomplishes this by reading timing constraints from the design.pcf physical constraints file.
Applications - Software
USING VARIABLES AND ATTRIBUTES
With Exemplar’s LeonardoSpectrum, you can easily control every aspect of your design.
by Tom Hill, Technical Marketing Manager, Exemplar Logic, [email protected]
L eonardoSpectrum is the FPGA Synthesis Tool from Exemplar Logic, used to Synthesize RTL
HDL code and target Xilinx Virtex devices.
Using Attributes and Variables
Argument Set_attribute Switch Description
Netlist object -port | -net | -instance Indicates the type of netlist object
Attribute Name -name Name of the attribute
Attribute Value -value Value of the attribute
LeonardoSpectrum’s optimization engine can be controlled globally (using global optimization variables) or at the netlist level (using attributes).
Global optimization variables affect the optimization of every block in the design. The lut_max_fanout variable provides a good example. If this is set to a value of 12 then all nets will be optimized with a fanout no greater than 12.
Table 1 - set_attribute command arguments.
device, which are primarily used to drive clock lines.
Often, all of these cells are not required for primary clocks. LeonardoSpectrum will automatically identify high fanout internal clocks and insert all unused global buffers into those nets. This functionality is enabled by default and controlled by the global variable:
> set insert_bufs_for_internal_clock true
Attributes provide a way to make a modification to a specific net, cell, or instance. An attribute is a property assigned to an object, which affects the optimization of only that object. You can set attributes prior to RTL synthesis by using a standard VHDL attribute statement or Verilog attribute comment statement as follows:
There may be situations where you want to force a BUFG cell onto a net or prevent the automatic buffering of a high-fanout net. Attributes can be used in both of these cases to control optimization.
Use the following command to force a BUFG cell into an internal net:
> set_attribute netname -net -name PAD -value bufg
• Setting Attributes in Verilog:
//Exemplar attribute <signal name> <attribute name>
• Setting Attributes in VHDL:
ATTRIBUTE <attribute name> : <attribute VHDL type>;
ATTRIBUTE <attribute name> of <signal name> : signal is <attribute value>;
Use the following command to prevent automatic
BUFG insertion on particular net:
> Set_attribute netname -net -name NOBUF -value TRUE
Controlling Virtex Low Skew Lines
Attributes can also be set on objects after synthesis using the set_attribute command as follows: usage : “set_attribute” [<object_list>] <netlist object>
<attribute name> <value>
For example: set_attribute u1 -instance -name dont_touch -value TRUE
Controlling Clock Buffers
Virtex and VirtexE FPGAs contain 4 BUFG cells per
If high-fanout clock nets still exist after all the clock buffers have been exhausted, LeonardoSpectrum will use the Virtex secondary global lines to minimize skew by applying an attribute called “MAXSKEW” to each individual, high-fanout clock net. You can instruct LeonardoSpectrum to perform this operation automatically on the entire design by setting the system variable virtex_apply_maxskew to a specified
maximum skew value as follows:
> Set virtex_apply_maxskew 7
The “maxskew” attributes can be used to specify a maximum skew on a particular net. Use the following command to limit skew on a particular net:
> Set_attribute <internal netname> -net -name MAXSKEW
If you set a very low skew value, one that the
Xilinx Alliance Series software is unable to meet, then an error will be issued during place and route. For that reason it is important not to over constrain the skew value.
Fanout, which is usually controlled globally, is set to a default of 64 in LeonardoSpectrum, for Virtex devices.
This is generous but usually gives good results. If timing is difficult to meet in the place and route environment, especially if large routing delays are the culprit, then you can modify design fanout in certain areas.
Using the lut_max_fanout attribute, LeonardoSpectrum allows you to override a global fanout specification with a unique value for a particular block or net. Net fanout can be specified on a global basis using the following command:
> set lut_max_fanout 12
Use the following command to redefine the fanout on a specific net from 64 (default) to 16:
> Set_attribute mynetname -net -name lut_max_fanout -value 16
Controlling IOB Registers
LeonardoSpectrum, by default, does not optimize registers to the IOB. When loading the Virtex technology, you are presented with an option to “Map IOB
Registers,” which will set the global optimization variable:
> set virtex_map_iob_registers TRUE
Once set this variable will instruct
LeonardoSpectrum to pull all possible registers into the Virtex IOB blocks. When a single register is used to drive more than 1 output port LeonardoSpectrum will replicate that register once for each additional port. Often, however, users wish to pull only a few select registers into the IOB. This can be accomplished by assigning an attribute called “IOB” to the particular register instance.
Use the following command to force a register with an instance name called “reg_state(7)” into the
IOB: set_attribute reg_state(7) -instance -name IOB -value TRUE
You can also use wildcards “*” to set all the registers of a bus at once. For example:
Controlling Block RAM
Both global variables and attributes can be used to control block RAM inferencing. When developing a single module of a larger device you may wish to allocate all the block RAM resources to another block. In this case, using the global variable to disable the block
RAM inference makes the most sense. You can do this by using the following command:
> set extract_ram FALSE
When block RAM resources start to run low, you may choose to disable block RAM inferencing on a section of the design. You can accomplish this by assigning the block_ram attribute to the storage signal used during the RTL RAM inference. You can do this in the RTL code by setting an attribute on the memory signal.
Reg [7:0] mem[63:0]
//Exemplar attribute mem block_ram FALSE
TYPE mem_type IS ARRAY 0 TO 256 OF std_logic_vector
(15 DOWNTO 0);
SIGNAL mem : mem_type;
ATTRIBUTE block_ram : Boolean;
ATTRIBUTE block_ram of mem : signal is TRUE;
The block_ram attribute can also be applied to the
“mem” signal, after synthesis, from the
LeonardoSpectrum command line.
LeonardoSpectrum’s design browser can help you identify correct signal pathnames for hierarchical blocks. For flat designs the signal name alone would be sufficient. Below is an example of using the set_attribute command to disable RAM inferencing on a memory signal “mem” that resides within a subblock of the design “blockA:”
> Set_attribute .work.blockA.rtl.mem* -net -name block_ram value false
With LeonardoSpectrum, it’s easy to optimize Virtex
FPGA designs. For more information about
LeonardoSpectrum, see www.exemplar.com.
Spring Tide Uses Xilinx FPGAs for a New
The Virtex devices are already proving to our customers the timesaving advantages that these FPGAs bring to their design cycle.
by Tamara Snowden, Public Relations, Xilinx, [email protected]
T he phenomenal success of the Internet and the universal adoption of the Internet
Protocol (IP) are driving profound changes in the telecommunications industry. Spring Tide
Network’s IP Service Switch 5000 creates a new service layer for the public Internet Protocol network, and they chose eleven Xilinx FPGAs, including SpartanXL and Virtex devices, based upon the strong set of development tools and design flexibility.
on ASIC science projects only to come up empty and disappoint their investors.”
“Working with Xilinx was a very simple decision for us because Xilinx offered tremendous gate density and speed... As a result, we were able to leverage their product to achieve our design objectives...”
“We began working with Xilinx as soon as we assembled our team. As a start-up, time to market was our overriding concern. Working with Xilinx was a very simple decision for us because Xilinx offered tremendous gate density and speed,” said Steve Akers, co-founder and chief technology officer. “As a result, we were able to leverage their product to achieve our design objectives while avoiding the risk associated with an ambitious ASIC project.”
“For our company, it was critical that we get to market in a timely fashion. Embarking on an ambitious ASIC program would have introduced too much risk to our plan,” said Bob Sullebarger, vice president of marketing for Spring Tide.
“We’ve seen several companies make the big bet
About Spring Tide
Spring Tide switches are designed to create dedicated, secure tunnels between users, on demand, automatically setting up and tearing down point-to-point connections as needed.
They are the only carrier-class, “any-toany” data services tunnel switches capable of full service delivery at line speed. In addition, Spring
Tide’s hardware and software-
About Spring Tide
Spring Tide was founded by Stephen
Collins, former vice president of marketing at StarBurst Communications, and Steve
Akers, former CTO and vice president of based intelligence gives each flow the service it requires, such as filtering, tagging, classification, queuing, and address translation.
advanced technology development at Shiva
Corporation. Led by President and CEO
Allan Wallack, the company has assembled a highly experienced product development deal of design flexibility as we architected our product,” said
team with engineering talent drawn from
The Spring Tide switch’s architecture features dedicated processors for handling many of the leading networking and computer companies in the Boston area. For more information, visit the Spring Tide website at www.springtidenet.com.
the functions of flow classification and queuing; packet forwarding, routing and session management, encryption and compression, and key generation. A fully configured IP Service
Switch 5000 contains more than one hundred
The Virtex technology is the combination of leading-edge process technology, a systemlevel feature set, and breakthrough software technology.
To build a multi-million-gate
FPGA, specific design optimization of the technology was necessary. The Virtex architecture special and general-purpose processors, as well as hundreds of megabytes of memory for context represents aggressive use of an advanced 0.22micron process to pack millions of gates with full and state information.
utilization of five metal layers for an abundance of high performance routing tracks. The architec-
The 5000 processes as many as 100,000 simultaneous secure customer sessions in a service provider’s point of presence, without sacrificing Quality of Service or performance goals.
Competing switches can process and track only a small fraction of that number. The 5000 switches all flows at line speed, regardless of any value-added processing (such as encryption), with less than 50 microseconds latency. ture is also easily scalable from 50,000 to one million system gates to offer the robust feature set across a wide range of densities in nine devices.
“The Xilinx parts allowed us to optimize the performance of our system by allowing us to run at 66 MHz,” said Akers. “The overall design is very flexible and only Xilinx could give us a device capability supporting between 500,000 and 3.2 million gates.”
The IP Service Switch 5000 contains eleven
Xilinx devices; 500k Virtex devices are installed in two locations: as a front-end packet scheduler to feed streams to the dual PowerPC processors on the tunnel card, and as a scheduler for the security subsystem. “We went with Xilinx for several reasons. First, Xilinx offered the strongest set of development tools available.
Second, Xilinx had the highest device speed, which helps us from a performance perspective.
In addition, Xilinx products offered us a great
“The Virtex devices are already proving to our customers the time-saving advantages that these FPGAs bring to their design cycle. The density and performance levels, combined with the popular system-level features and industry-leading FPGA price points make Virtex-E FPGAs an even more compelling alternative to ASICs,” said
Wim Roelandts, Xilinx president and CEO.
VisiCom Uses Xilinx FPGAs for a
The combination of the Virtex hardware, associated software tools, and engineering process improvements have proven to be a great success.
by Tamara Snowden, Public Relations, Xilinx, [email protected]
W hen VisiCom needed a high capacity, high performance FPGA for a realtime video processing board, they chose the Virtex series FPGAs from Xilinx. Since
1997, VisiCom has specialized in using FPGAs to rapidly deploy customer-specific real-time image and video processing systems for a variety of applications, including semiconductor process equipment, medical imaging, scientific imaging, factory floor automation and defense/aerospace electronics.
VisiCom’s FPGA-based approach combines the real-time performance of hardware-accelerated algorithms with the ability to modify or upgrade the algorithm without physically changing the hardware. This provides a means to reconfigure hardware and software quickly and efficiently to suit specific customer requirements, resulting in reduced time to market, significant cost savings, and better lifecycle support. In addition, the use of FPGA technology avoids the non-recurring engineering costs usually associated with fabricating ASICs, while still providing most, if not all, of the performance features.
“With the FPGAs used in our earlier projects, we found we were spending a great deal of time
‘hand-crafting’ algorithms to make them fit in the part and achieve timing closure,” stated Ron
Hawkins, chief operating officer of VisiCom’s
Computer Products Division. “A lot of effort from our most experienced engineers was required, tying up valuable engineering resources and delaying customer deliveries. We realized this was unacceptable to us and our customers, and something had to change.” VisiCom’s engineers needed a solution that would allow them to spend more time on design and simulation and less on post route and hardware debug.
Figure 1 - Visicom board.
The opportunity to change occurred when
VisiCom was selected to design a custom embedded, reconfigurable image processing module for a medical imaging system. The system in question had demanding requirements; it called for processing full bandwidth video
(“4:4:4” luminance and chrominance bands) in real-time using several image processing and ancillary operations. The processing requirements included color space conversion, enhancement (convolution), and affine transformation (rotation, scaling, and translation). These
Using the Virtex Series
VisiCom’s engineers chose the largest capacity
FPGA available from Xilinx, the million-gate
Virtex series, as their solution. Virtex provides more gates and routing resources, making it easier to achieve timing closure with the synthesis tools available. “We bet on two things,” continued Hawkins.
VisiCom’s FPGA-based approach....provides a means to reconfigure hardware and software quickly and efficiently to suit specific customer requirements, resulting in reduced time to market, significant cost savings, and better lifecycle support.
operations represented a significantly higher degree of integration on silicon than had been previously attempted by VisiCom engineers. In addition, the new module was replacing an existing module that had become obsolete. The new module needed to accommodate specific size and power constraints.
“First, we bet that Moore’s Law would apply to FPGAs, that equivalent gate counts would continue to increase rapidly, with the cost per gate declining over time. Therefore, we concluded that even if the target part did not meet our cost requirements, it would by the time we fielded the product. This meant we could select a part that had plenty of resources to meet our needs and still have confidence that our cost targets in production could be met.”
“Second, we bet that the EDA tool vendors would continue to improve their products to achieve optimum utilization and routing within
FPGAs. We decided that the problem of optimally synthesizing, placing and routing logic was a problem for the tool vendors to solve, not us.
This assumption meant we could devote a greater percentage of effort to analyzing our application problem and designing a solution, and spend less time getting it to work in the part. We relied on the EDA tools to handle this, permitting better use of our engineering resources to focus on solving the application problem at hand.”
These assumptions proved correct. VisiCom engineers converted from schematic design
capture to VHSIC Hardware Description
Language (VHDL). The imaging algorithm designs had to be more abstract, modular, and reusable, to meet time-to-market constraints and support reuse in future designs. The softwarelike properties afforded by VHDL, such as encapsulation and parameterization, offered significant advantages relative to schematic capture.
The support in VHDL for the creation of design libraries and reusable components was also essential to VisiCom’s strategy for rapidly fielding future FPGA-based imaging systems to meet specific customer requirements.
Another benefit of using the Virtex series was the high number of usable I/O pins. This feature permitted VisiCom engineers to lock the
FPGA pinout early in the project, so the physical board design and fabrication could proceed concurrently with the VHDL coding of the imaging algorithms. The end result was the imaging board was ready and waiting upon completion of the fully simulated imaging algorithm. “The engineers downloaded the bitstream, made a few tweaks, and our imaging board was up and running,” said Hawkins.
The company’s engineers also placed more emphasis on establishing a formal design methodology to guide the work effort from analysis and high level design of the imaging algorithm, through VHDL coding, and ultimately, synthesis into the Virtex part. The software tools available for the Xilinx Virtex series complemented this methodology and the use of VHDL.
VisiCom integrated a high-level signal processing and analysis tool, MATLAB(tm) from
MathWorks, into the design flow, creating a framework for analyzing and verifying the imaging algorithms before beginning VHDL coding.
The combination of the Virtex hardware, associated software tools, and engineering process improvements have proven a great success in creating a solid product for use in the customer’s medical imaging system. VisiCom not only achieved a faster time-to-market, but also completed a real-time imaging project of significantly greater scope than had been previously attempted. As a basis for comparison, VisiCom engineers estimated that seven commercial VLSI imaging chips and SRAM memory would have been required to achieve the same results at a greater cost than VisiCom’s solution, which used a single Xilinx XCV-200 part with SDRAM memory.
Extensive use was made of VHDL test benches for thorough simulation and verification of the design after it was committed to VHDL. A procedure was established for processing video fields with both the MATLAB and VHDL simulations, and comparing the results to ensure proper implementation. The methodology also includes use of post-route simulation to ensure timing closure. As a result of using MATLAB,
VHDL, and extensive simulation, relatively little time was spent debugging the physical hardware.
In the customer’s medical imaging system, a single Virtex part replaced multiple smaller capacity parts and provided a smaller footprint, lower power requirements, more capability, and lower overall cost. Based on their experiences with this project, VisiCom’s design team is now working on their next real-time imaging design using the Virtex-E architecture.
For more information, visit the VisiCom website at: www.visicom.com. Or, e-mail: [email protected]
Electronic Distribution is Changing with the Times
Distributors continue to add new products and services to meet your needs.
by Russ Sinagra, Distribution Marketing Manager HVG, [email protected]
T he market demands for shorter development cycles have had an effect on every aspect of the industry, and distribution is no exception. Over the past ten years distributors have made significant investments in people and equipment in an effort to increase their value to you. Their services have ranged from simple selection and sorting processes (for special tolerance parts), to building custom cable assemblies, to creating very sophisticated worldwide logistics programs. However, there are new and growing concerns related to today’s global economy, and distributors are doing all they can to address these concerns.
Distributors Become an Important Resource
Distributors still make money the old fashioned way, by selling components, but they also realize they must now provide design services for their customers as well. Thus, they have invested millions of dollars over the past years for facilities, state of the art software, support engineering staff, and Intellectual Property (IP) libraries, in an effort to fill the design needs of their customers.
Xilinx is very fortunate because we have three distributors that offer various levels of design services and support: Avnet Design
Services, Memec Design Services, and
NuHorizons Electronics. Their services include on site customer seminars, Web-based seminars, training, and full design support from concept to complete turnkey solutions. Plus, they have developed their own IP, reference designs, demo boards, and software kits to aid you with your designs.
Engineering Concerns for the Next
Because device and software technology has changed dramatically over the past two years it has been difficult for many customers to keep up. For example, during the “Design Con 2000” show in Santa Clara, California (last February), attendees were asked to list their primary concerns for the next three years; the following are the top four:
• Getting their product to market ahead of the competition.
• Keeping up with current market demands.
• Obtaining needed technical expertise.
• Reducing inventory liability.
When asked how they were planning to address these concerns, the majority answered:
The Xilinx distributors have taken a very active role in helping you keep product costs and development time to a minimum. If you would like to know more about their services, visit them on the Web at:
* Avnet Design Services: http://www.ads.avnet.com/
* Memec Design Services: www.memecdesign.com/
* NuHorizons Electronics: http://www.nuhorizons.com/
Column - Trade Shows
E V E R Y W H E R E
10 Million Gates
See Booth #3629 at DAC in Los Angeles, June 5-7th and see just how hot Xilinx is!
by Darby Mason-Merchant, Trade Show Manager, Xilinx,
Inc., [email protected]
A t the 37th Design Automation
Conference, Club Xilinx will be the hot spot where you can see why FPGAs and
CPLDs are now center stage in digital design.
Stop by and see our New Foundation ISE and
3.1i software; an entire suite of design tools delivering the first design environment for our new 10 million gate FPGAs. Also, check out the
IP Center for everything you need for design reuse, and catch up on the amazing new developments in Internet Reconfigurable Logic— upgradability via the Internet.
Year 2000 Worldwide Xilinx Trade Show Schedules
Year 2000 North American Trade Show Schedule
Embedded Computing Shows 2000
AC Developers Conference 2000
37th Design Automation Conference
2000 ASEE Conference and Expo
WITI Technology Summit 2000
Embedded Internet Conference 2000
NCF / InfoVision 2000
Frontiers in Education 2000
Santa Clara, CA
Los Angeles, CA
St. Louis, MO
Santa Clara, CA
San Jose, CA
Kansas City, MI
From satellites to cellular, networks to MP3s, routers to cable modems, Xilinx is everywhere.
To find out more about Club Xilinx at DAC, go to: www.xilinx.com. For more information about
Xilinx Worldwide Trade Show Programs, please contact one of the following Xilinx team members or see our website at: http://www.xilinx.com/company/tradeshows.htm
• European Shows: Andrea Fionda at: [email protected]
• Japanese Shows: Renji Mikami at: [email protected]
• SouthEast Asian Shows: Mary Leung at: [email protected]
Year 2000 European Trade Show Schedule
Nov 2000 Electronica 2000
Year 2000 South East Asian Trade Show Schedule
Year 2000 Japanese Trade Show Schedule
Jan 27-28 EDA Techno Fair 2000
Nov 2000 Micon System Tool Fair 2000
Column - Customer Education
Xilinx announces a new Advanced FPGA class.
By Renne Ricciardi, [email protected]
X ilinx Customer Education offers a number of public classes on-site training courses and on-line E-Learning modules to help you quickly learn the latest development tools.
Our public courses are held at various locations around the world, and our E-Learning modules are available online, via the Web. On-site classes are scheduled on request.
Online E-Learning Modules
Each E-Learning module is a one-hour session that focuses on specific product solutions, and you have the ability to ask questions of the instructor, get answers, and collaborate with other students. Our E-Learning modules are available now as live, on-line sessions ($100.00
USD per module) and will soon be available in recorded versions ($70.00 USD per module). New modules are added frequently, so check our website to see the latest sessions and schedules.
New On-site Advanced FPGA Class
Our new on-site, two-day Advanced FPGA class is designed to help you:
• Get higher performance from your design.
• Optimize your code for the Virtex architecture.
• Take advantage of the more advanced capabilities of the implementation tools, such as
Multi-pass Place and Route, and re-entrant routing.
Course topics include:
• Targeting Virtex-specific features with
Synopsys FPGA Express and Synplicity
• Using advanced timing constraints.
• Using the FPGA Editor.
• Using Xflow.
• Creating Virtex memory applications.
Class size is limited, so enroll early. To learn more about our classes or to register you can contact the registrar at 877-XLX-CLASS (877-
959-2527), or go to: http://www.support.xilinx.
Advanced FPGA Class Schedule
May 9 & 10
June 8 & 9
June 13 & 14
July 12 & 13
July 26 & 27
Aug 9 & 10
August 24 & 25
San Jose, California
San Jose, California
San Jose, California
San Jose, California
Alliance 3.1 Update
Intro to Verilog San Jose
Foundation ISE Design Entry San Jose
FPGA Design-ASIC User
Alliance 3.1 Update
Introduction to VHDL
E ach Virtex family has its own unique features to meet different application requirements. All devices have both distributed
RAM and block RAM, and between four and eight DLLs for efficient clock management.
• The Virtex family, consisting of devices that range from 50K up to 1 million logic gates, supports 17 I/O standards, and offers 5V PCI compliance.
• The Virtex-E family offers the highest logic gate count available for any FPGA, ranging from 50K up to 3.2 million system gates, and supports 20 I/O standards including LVPECL,
LVDS, and Bus LVDS differential signaling.
• The Virtex-E Extended Memory family consists of two devices that have a high RAMto-logic gate ratio that is targeted for specific applications such as gigabit per second network switches and high definition graphics.
The XC4000XLA family is part of the broad spectrum of Xilinx XC4000 FPGA series which offers the broadest choice of 5V, 3.3V, and 2.5V
devices. The XC4000XLA family consists of eight members ranging from 30,000 to 180,000 system gates and features patented SelectRAM distributed RAM that can be used as single-port or dual-port memory.
See www.xilinx.com/__________________? for more information.
FPGA Product Selection Matrix
FPGA Product Selection Matrix
DEVICES KEY FEATURES
* I/Os are 5V tolerant
** 5 Volt tolerant I/Os with external resistor
X = Core and I/O voltage
I/Os = I/O voltage supported
576 1536 192 12/24 Y
784 2016 205 12/24 Y
1024 2560 256 12/24 Y
1296 3168 288 12/24 Y
1600 3840 320 12/24 Y
1936 4576 352 12/24 Y
2304 5376 384 12/24 Y
3136 7168 448 12/24 Y
384 1536 180 2/24 Y
600 2400 180 2/24 Y
864 3456 260 2/24 Y
1176 4704 284 2/24 Y
1536 6144 316 2/24 Y
21168 254K 511K-888K 406K 56x84
27648 332K 622K-1,124K 512K 64x96
1728 21K 47K-72K 88K 16x24
10800 130K 1,068K-1,307K 710K 40x60
15552 187K 679K-986K 504K 48x72
21168 254K 2,569K-3,062K 1414K 56x84
27648 332K 1,146K-1,569K 768K 64x96
6144 24576 512 2/24 Y
384 1536 176 2/24 Y
34992 420K 1,628K-2,189K 1062K 72x108 7776 31104 724 2/24 Y
43200 518K 1,857K-2,542K 1240K 80x120 9600 38400 804 2/24 Y
57132 686K 2,221K-3,264K 1530K 92x138 12696 50784 804 2/24 Y
73008 876K 2,608K-4,074K 1846K 104x156 16224 64896 804 2/24 Y
– – – X
– – – X
– – – X
– – – X
– – – X
– – X
– – X
– – X
– – X
– – X
– X I/O *
– X I/O *
– X I/O *
– X I/O *
– X I/O *
– – X
– – X
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
X I/O I/O **
S ay hello to a new level of performance; the
Spartan-II family now includes devices with over 200,000 system gates, and you get 100,000 system gates for under $10, at speeds of 200Mhz and beyond, giving you design flexibility that’s hard to beat. These low-powered, 2.5-V devices feature I/Os that operate at up to 3.3V with full 5-V tolerance. Spartan-II devices also feature multiple Delay Locked
Loops, on-chip RAM (block and distributed), and versatile I/O technology that supports over 16 high-performance interface standards. You get all this in an FPGA that offers unlimited programmability, and can even be upgraded in the field, remotely, over any network.
Robust Feature Set
• Flexible on-chip distributed and block memory.
• Four digital Delay Locked Loops for efficient chip-level/board-level clock management.
• Select I/O Technology for interfacing with all major bus standards such as HSTL, GTL,
SSTL, and so on.
• Full PCI compliance.
• System speeds over 200 MHz.
• Power management.
Extensive Design Support
• Complete suite of design tools.
• Extensive core support.
• Compile designs in minutes.
Advantages Over ASICs
• No costly NRE charges.
• No time consuming vector generation needed.
• All devices are 100% tested by Xilinx.
• Field upgradeable (remotely upgradeable, using Xilinx Online technology).
• No lengthy prototype or production lead times.
• Priced aggressively against comparable
For more information see www.xilinx.com/products/spartan2
FPGA Product Selection Matrix
FPGA Product Selection Matrix
DEVICES KEY FEATURES
* I/Os are tolerant
X = Core and I/O voltage
I/Os = I/O voltage supported
400 1120 160
576 1536 192
784 2016 205
360 77 12/24 Y
784 2016 224 12/24 Y
96 384 86 2/24 Y
384 1536 176 2/24 Y
600 2400 196 2/24 Y
102K 24x36 864 3456 260 2/24 Y
130K 28x42 1,176 4704 284 2/24 Y
XC9500 and CoolRunner CPLDs
X ilinx WebPOWERED Software Solutions offer yhou the flexibiity to target the
XC9500 and CoolRunner Series CPLDs on-line or on your desktop.
WebFITTER is an on-line device fitting and evaluation tool which accepts HDL, ABEL, or netlist files and provides all reports, simulation models, and programming files.
WebPACK downloadable desktop solutions offer free CPLD software modules for ABEL and
HDL synthesis, HDL simulation, device fitting, and JTAG programming.
See www.xilinx.com/products/software/webpowered.htm for more information.
Ultra Low Power
Ultra Low Power
Ultra Low Power
Ultra Low Power
* Philips part number
QML Certified FPGAs
X ilinx is the leading suppler of High-
Reliability programmable logic devices to the aerospace and defense markets.
These devices are used in a wide range of applications such as electronic warfare, missile guidance and targeting, RADAR, SONAR, communications, signal processing, avionics, and satellites. The Xilinx QPRO family of ceramic and plastic QML products (Qualified Manufacturers
Listing), certified to MIL-PRF-38585, provide you with advanced programmable logic solutions for next generation designs, and include select products that are radiation hardened for use in satellite and other space applications.
The Xilinx QPRO family addresses the issues that are critical to the aerospace and defense market:
• QML/Best commercial practices. Commercial manufacturing strengths result in more efficient process flows.
• Performance-based solutions, including costeffective plastic packages.
• Reliability of supply. Controlled mask sets and processes insure the same quality devices, every time, without variation, which remain in production for an extended time.
• Off-the-shelf ASIC solutions. Standard devices readilly available, no need for custom logic and gate arrays.
Being certified to MIL-PRF-35835 QML, complemented by ISO-9000 certification, results in an overall product quality platform that makes
Xilinx a world-class supplier of programmable logic devices. You can confidently design with
Xilinx for High-Reliability systems with the assurance that you are getting unsurpassed quality and reliability, and a long-term commitment to the aerospace and defense market.
See www.xilinx.com/products/hirel_qml.htm for more information.
QPRO QML-Certified PROMs
Density DD8 SO20 CC44 PC44
* XQR devices are Radiation Hardened.
** XQ devices only.
FPGA Product Selection Matrix
FPGA Product Selection Matrix
DEVICES KEY FEATURES
SelectI/O 4 DLLs
* I/Os are tolerant
** XQR and XQVR devices are Radiation Hardened
X = Core and I/O voltage
I/Os = I/O voltage supported
18K 24x24 576 1536 192 12/24 Y
42K 36x36 1296 3168 288 12/24 Y
74K 48x48 2304 5376 384 12/24 Y
100K 56x56 3136 7168 448 12/24 Y
2700 32K 72K-109K 78K 20x30 600 2400 180 2/24 Y
27648 332K 622K-1,124K 512K 64x96 6144 24576 512 2/24 Y
– – X
– – X
– – X
– – X
X I/O *
X I/O *
X I/O *
X I/O *
Solutions on the
O n the Xilinx website you can find anything you need for configuring Xilinx
PLDs. Here are some examples of what’s available:
• Downloadable Java files that allow for crossproduct and multi-platform support for In-
System-Programming (ISP). http://www.xilinx.com/products/software/sx/javapi.htm
• Downloadable JTAG tools (from WebPACK). http://www.xilinx.com/isp/toolbox.htm
• Download cable information (Xilinx Parallel and High-Performance MultiLINX USB cable). http://www.xilinx.com/isp/toolbox.htm
• Design considerations for In-System-
• Embedded micro-controller support for performing JTAG operations, and using JTAG serial vector format (SVF) files for embedded processors. http://www.xilinx.com/isp/ess.htm
• Boundary Scan tutorials and information on the new IEEE1532 specification which enables you to generate programming algorithms using only BSDL files.
• Machine level language translators for
Genrad, H-P, and Teradyne among others. http://www.xilinx.com/isp/ate.htm
• Information on the industry leading programmers from BP Microsystems and Data I/O along with many other popular models. Also covers the Xilinx HW130 programmer and device adapters. ttp://www.xilinx.com/support/programr/ps.htm
• Boundary Scan tools from six of the industryleading tool suppliers, with details on the
Xilinx device certification program as well as details of each tool that can be used for test and debug as well as programming.
• Order cables on line. Compare the features of each of the cables Xilinx has to offer, and choose the cable that meets your needs.
• Register here for all the latest information on any programming algorithm changes Xilinx makes to non-volatile (CPLD and PROM) product offerings.
• Overview of Xilinx PROMS (one time (OTP) and In-System-Programmable).
See www.xilinx.com/isp/isp.htm for complete information on all Xilinx Configuration Solutions.
X ilinx offers a full range of configuration memories optimized for use with Xilinx
FPGAs. Our in-system re-programmable
XC18V00 family provides a feature rich, fast configuration solution unmatched by any other configuration PROM available today, and provides a cost-effective method for reprogramming and storing large Xilinx FPGA bitstreams. Our lowcost XC1700 and XC17S00 families are an ideal configuration solution for cost sensitive applications.
Our PROM product lines are designed to meet the same stringent demands as our high performance FPGAs and CPLDs, taking full advantage of our advanced processing technologies. They were developed in close cooperation with Xilinx FPGA designers for optimal performance and reliability.
See www.xilinx.com/products/configsolu.htm or more information
3.3V Configuration PROMs
Device Density PD8 SO20 PC20 PC44 VQ44
XC1765EL(X) 64Kb X X X X
XC17128EL(X) 128Kb X X X
XC17256EL(X) 256Kb X X X
XC17512L 512Kb X X X
XC1701L 1Mb X X X
XC1702L 2Mb X X
XC18V128 128Kb X X X
XC18V512 512Kb X X X X
XC18V01 1Mb X X X X
XC18V02 2Mb X X X
XC18V04 4Mb X X X
Note: XC1700EL parts are marked with an “X” instead of “EL”
3.3V Configuration PROMs for Spartan-XL/Spartan-II
Device Configuration Bits PROM Solution PD8 VO8 SO20
XCS05XL 54,544 XC17S05XL X X
XCS10XL 95,752 XC17S10XL X X
XC2S15 197,696 XC17S15XL X X
XCS20XL 179,160 XC17S20XL X X
XCS30XL 249,168 XC17S30XL X X
XC2S30 336,768 XC17S30XL X X
XCS40XL 330,696 XC17S40XL X X* X
XC2S50 559,232 XC17S50XL X X
XC2S100 781,248 XC17S100XL X X
XC2S150 1,041,128 XC17S150XL X X
XC2S200 1,335,872 XC17S200XL X X
* In development
Configuration PROMs for Virtex-E
PD8 PC20 SO20 PC44 VQ44
XCV50E 630,048 01 X** X X X***
XCV100E 863,840 01 X** X X X***
XCV200E 1,442,106 02 X X
XCV300E 1,875,648 02 X X
XCV400E 2,693,440 04 X X
XCV405E 3,430,400 04 X X
XCV600E 3,961,632 04 X X
XCV812E 6,519,648 04+04 or 08* X X
XCV1000E 6,587,520 04+04 or 08* X
XCV1600E 8,308,992 04+04 or 08* X
XCV2000E 10,159,648 04+08* or 16* X X
XCV2600E 12,922,336 16* X X
XCV3200E 16,283,712 16* X X
* In development
** Available on XC17xx only
*** Available in XC18Vxx only
Configuration PROMs for Virtex
PD8 PC20 SO20 PC44 VQ44
XCV50 558,048 01 X* X X X**
XCV100 780,064 01 X* X X X**
XCV150 1,038,944 01 X* X X X**
XCV200 1,334,688 02 X X
XCV300 1,750,656 02 X X
XCV400 2,544,896 04 X X
XCV600 3,606,816 04 X X
XCV800 4,714,400 04 + 512 X X
XCV1000 6,126,528 04 + 02 X X
* Available on XC17xx only
** Available on XC18Vxx only
hy limit yourself to choosing between a DSP processor or an ASIC when you can have the best of both worlds? Like the world of DSP processors, Xilinx ® gives you the flexibility of reconfigurable off-the-shelf devices. Like the world of ASICs, we also give you high performance in a single chip.
provides the IP cores and tools you need to take you from definition to implementation. Build your system using our LogiCORE
™ products — ranging from parameterizable adders and multipliers to FIR, DDS, FFT, and FEC. Use our GUI-based CORE Generator ™ system to automatically generate optimal implementations, and use our new MATLAB SIMULINK interface for system simulation and HDL generation.
To complement the Xilinx solution, you have access to our AllianceCORE ™ and XPERTS ™ partners providing additional IP cores, prototyping tools, and design services.
All enabled by our powerhouse Virtex ™ , Virtex-E, and Spartan ™ -II product families, it’s a total solution that guarantees your time to market success in bringing your product to the world.
The Programmable Logic Company
© 2000, Xilinx, Inc. All rights reserved. The Xilinx name and the Xilinx logo are registered trademarks, AllianceCORE, Core Generator, LogiCORE, Spartan, Virtex, and XPERTS are trademarks, and The Programmable Logic Company is a service mark of Xilinx, Inc. All other trademarks and registered trademarks are the property of their respective owners.
2100 Logic Drive
San Jose, CA 95124-3450
First Class Presort
Permit No. 2196
San Jose, CA
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project