WP462 – UltraScale Devices Maximize Design Integrity with Industry

WP462 – UltraScale Devices Maximize Design Integrity with Industry
White Paper: UltraScale Devices
WP462 (v1.0) February 26, 2015
UltraScale Devices Maximize Design
Integrity with Industry-Leading
SEU Resilience and Mitigation
By: Derek Curd and Eric Crabill
UltraScale™ architecture provides extraordinary SEU
resilience and mitigation through its industry-leading
offering of robust soft-error solutions.
ABSTRACT
Xilinx introduced the UltraScale architecture, the industry's f irst ASIC-class fully
programmable architecture, to enable ultra-high performance applications
beyond the capabilities of prior generations and competing programmable
solutions.
• Rapid process technology scaling and architectural innovations allow
UltraScale devices to achieve levels of system performance, capacity, and
power eff iciency that make FPGAs an increasingly compelling alternative to
ASICs. As such, UltraScale FPGAs can be expected to be found at the heart
of systems that demand high availability, high reliability, and adherence to
the strictest functional safety requirements.
• To maximize the integrity of UltraScale architecture-based designs,
Kintex® UltraScale and Virtex® UltraScale devices offer industry-leading
resilience in the presence of single-event upsets (SEUs) through product
innovations that, in many applications, reduce or eliminate the need for
additional soft-error mitigation solutions. UltraScale devices exhibit up to
3X lower SEU failure-in-time (FIT) per Mb and 2X faster detection and
correction of soft errors than prior-generation Xilinx devices.
• For systems that require the highest availability and reliability, Xilinx
complements its UltraScale architecture innovations with analysis tools, IP,
design techniques, and verif ication flows.
Xilinx stands as the industry leader with its offering of the complete package —
the most robust, flexible, and comprehensive SEU mitigation solution available.
© Copyright 2015 Xilinx, Inc. Xilinx, the Xilinx logo, Artix, ISE, Kintex, Spartan, Virtex, Vivado, Zynq, and other designated brands included herein are trademarks of Xilinx
in the United States and other countries. All other trademarks are the property of their respective owners.
WP462 (v1.0) February 26, 2015
www.xilinx.com
1
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Introduction
The benefits of advanced process technologies for increasing device capacity, system performance,
system integration, and power efficiency cannot be overstated. FPGAs, in particular, are a
beneficiary of the ongoing migration to ever smaller process geometries. At the 20 nm node,
process scaling, coupled with major architectural innovations, has enabled Xilinx UltraScale FPGAs
to deliver unprecedented ASIC-class capabilities in a programmable device. Meanwhile, traditional
ASIC solutions continue to struggle with escalating development costs and the risks of committing
to a fixed solution in an increasingly dynamic, rapidly changing marketplace.
Along with the many benefits of process scaling, shrinking geometries require increased attention
to mitigating various effects that, left unchecked, work to reduce design integrity. Single-event
upsets (SEUs) represent one type of effect that requires ongoing diligence and increased levels of
innovation to manage in advanced process technologies. SEUs are caused by ionizing radiation
sources that send charged particles through transistor junction regions, causing a change in the
state of storage elements such as memory cells.
If specific efforts are not made to reduce susceptibility to SEUs, an increase in the device’s intrinsic
soft-error rate as a consequence of process technology scaling is expected. While the transistor
cross-section that is sensitive to charge injection is reduced with process scaling, the charge
required to change the state of a storage element (Q CRIT) is also reduced, and the density of
susceptible elements (e.g., configuration memory cells) tends to increase significantly.
Xilinx has multiple generations of experience in design and layout techniques to reverse this trend,
with lower SEU FIT/Mb in each successive process node. For instance, as shown in Figure 1, without
additional innovation at the 20 nm node, the expectation would be a 2X increase in soft error rate
per megabit as compared to the 28 nm node. The UltraScale architecture utilizes over
40 proprietary, patented circuit design and layout techniques to reduce the intrinsic SEU FIT/Mb of
the device’s configuration memory. As a result, UltraScale devices are targeted to have up to 3X
lower SEU FIT/Mb than 28 nm devices, a 6X improvement relative to the expected path for the
20 nm node without such innovation.
X-Ref Target - Figure 1
Soft Error Rate (per Mb)
Xilinx FPGA Soft Error Rate Trend
Without innovation,
20 nm was expected to
increase 2X vs. 28 nm
UltraScale architecture:
• 3X lower than 28 nm
• 6X lower than 20 nm
without innovation
65 nm
40 nm
28 nm
20 nm
16 nm
WP462_01_020915
Figure 1: Xilinx FPGA Soft Error Rates vs. Process Technology Node
WP462 (v1.0) February 26, 2015
www.xilinx.com
2
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Single-Event Upsets in FPGAs
As shown in Figure 2, SEUs represent one category of soft error-related events that are part of a
broader classification of single-event effects (SEEs). Unlike competing solutions, Xilinx has a proven
and public record, rigorously validated by 3rd parties, of immunity to hard errors; consequently,
those effects are not considered here. With regard to soft errors (fully correctable, nondestructive
events), the main sources for terrestrial applications originate from indirect ionization, such as
neutron-induced upsets from the environment, and direct ionization, such as alpha
particle-induced upsets from contaminants in device packaging.
While all forms of soft errors in Xilinx devices are extremely rare, the contribution of single-event
transients (SETs) and single-event functional interrupts (SEFIs) to the device SEU FIT is very low. This
is attributable to the quantity, design, and layout of user-accessible storage elements, such as CLB
flip-flops, DSP registers, and I/O registers. Therefore, SEUs that impact the device configuration
RAM (and, to some extent, the on-chip block RAM memory elements) are the primary sources of
soft errors that require mitigation in Xilinx FPGAs.
X-Ref Target - Figure 2
SEE
(Single Event Effect)
Soft Error
Hard Error
SEFI
SET
SEU
(Single Event Transient)
(Single Event Upset)
(Single Event
Functional Interrupt)
SBU
MBU
(Single Bit Upset)
(Multiple Bit Upset)
SEL
SEB
(Single Event Latch-Up)
(Single Event Burnout)
Recoverable Errors
SEGR
(Single Event Gate
Rupture)
Non-Recoverable Errors
WP462_02_020915
Figure 2: Single-Event Effect Classifications
Applications Requiring SEU Mitigation
The significant intrinsic SEU FIT/Mb reduction for UltraScale FPGAs expands the already broad
application space that can benefit from these improvements without the need for additional SEU
mitigation efforts. For example, a fully utilized Kintex KU040 device can be expected to experience
a soft error that impacts design functionality only once every 325 years or so (benchmark location:
New York).
UltraScale devices are designed to be at the heart of systems requiring high availability, high
reliability, and functional safety. Some examples of such applications are shown in Table 1. For
these areas, Xilinx offers the most robust, flexible, and comprehensive solutions available when
additional SEU mitigation is required.
WP462 (v1.0) February 26, 2015
www.xilinx.com
3
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Table 1: Example Applications Where Additional SEU Mitigation Should Be Considered
High Availability
Wired communications
Wireless communications
High Reliability
Functional Safety
Servers
Medical
Data Storage
Avionics
SEU Solution Overview
Figure 3 illustrates Xilinx's multi-layered approach to SEU resilience and mitigation. The
comprehensive Xilinx UltraScale architecture SEU management solution is unmatched by ASICs or
competing FPGA offerings.
The three primary objectives of the UltraScale architecture SEU solution are:
•
Lower the intrinsic SEU FIT/Mb for devices through innovative silicon and packaging
techniques.
•
Provide additional design-specific SEU mitigation, when needed, through readily available IP,
integrated tool flows, and dedicated design techniques.
•
Enable design-specific analysis and verification to validate the mitigation solution.
X-Ref Target - Figure 3
is & Verificati
on
alys
n
A
Techniqu
sign
es
De
SEM IP
ckaging
Pa
Silicon
WP462_02_020915
Figure 3: Multi-layered SEU Resilience and Mitigation Solutions for UltraScale Devices
The last objective is of particular interest when evaluating other device options, such as ASICs. The
unique programmable structure of Xilinx FPGAs enables testing and verification of the device and
system behavior due to SEUs—which is not feasible in an ASIC solution. Techniques such as
targeted fault injection into critical circuits to characterize functional impact allows for the
development and testing of appropriate system responses to soft errors.
Given that soft errors cannot be eliminated in any device, understanding how the system responds
and having the ability to efficiently develop system-level mitigation methodologies is just as
important as reducing susceptibility to SEUs.
WP462 (v1.0) February 26, 2015
www.xilinx.com
4
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Increased Reliability through Innovation
To deliver on the objective of significantly lowering the intrinsic SEU FIT/Mb for UltraScale devices,
Xilinx has implemented multiple innovative solutions in circuit design, physical device layout, and
materials management to deliver FPGAs with 3X lower SEU FIT/Mb. The foundation of these efforts
starts with silicon innovations and continues with deliberate processing and packaging methods to
deliver a complete physical product with industry-leading SEU resilience.
Silicon
To counteract the tendency of SEU susceptibility to increase with shrinking process geometries, the
UltraScale architecture configuration RAM cell (the largest contributor to device SEU FIT) is
implemented with over 40 circuit and layout improvements to substantially increase its Q CRIT value.
In other words, the configuration RAM cell is designed and laid out to be very difficult to flip in the
event of a particle strike. Xilinx uses silicon test vehicles with configuration RAM array structures to
validate predicted simulation results before actual device implementation.
Furthermore, configuration RAM cells are interleaved to ensure that physically adjacent cells are
not contained in the same configuration frame. The device's built-in error detection and correction
logic calculates ECC values on a frame-by-frame basis. Interleaving the configuration RAM cells in
this manner significantly reduces the likelihood of a multiple-bit upset occurring in a single
configuration frame as the result of one SEU event. This, in turn, increases the effectiveness of the
error detection and correction logic such that UltraScale devices will be able to correct over 99.9%
of SEUs, whether they result in single- or multiple-bit upsets.
Finally, Xilinx devices typically have about 40% fewer configuration RAM cells compared to
competing FPGAs of equivalent density, resulting in an inherently lower device SEU FIT beyond the
silicon benefits already described.
As the second most significant contributor to total device SEU FIT, user memory cells in the device’s
block RAM also benefit from SEU mitigation. As with the configuration RAM, Xilinx has been
improving the intrinsic SEU FIT/Mb of the block RAM with each successive process node through
innovative circuit design and layout techniques. In addition, every block RAM has built-in
single-error correct, double-error detect (SECDED) logic that can be enabled through a parameter
associated with each block RAM primitive.
Packaging
To significantly reduce alpha particle-induced upsets from contaminants in device packaging,
UltraScale devices use only ultra-low alpha (ULA) packaging with strict controls on all materials
used in the assembly process. Materials used in the package underfill, microbumps, C4 bumps, and
molding compound are carefully specified and monitored to minimize alpha particle contributions
to the device SEU FIT.
WP462 (v1.0) February 26, 2015
www.xilinx.com
5
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Increased Availability through Design Methodology
While all applications benefit from the significantly lower intrinsic SEU FIT/Mb of UltraScale
devices, some applications have reliability, availability, or functional safety requirements that need
additional SEU mitigation. Xilinx meets the objective of providing additional mitigation for such
applications through built-in device capabilities, IP, and design techniques.
Figure 4 shows how the intrinsic SEU FIT/Mb of UltraScale devices already contributes to very high
availability by minimizing total system downtime. This example assumes that 100k units of a Kintex
KU040 device have been deployed in systems at sea level. By applying additional elements of the
UltraScale architecture SEU mitigation solution to such a design, system availability can be further
increased to meet the requirements of even the most demanding applications.
450
100.0000%
400
99.9998%
350
99.996%
99.9994%
300
99.9992%
250
99.9990%
200
99.9988%
150
Availability
Down Time (seconds/yr)
X-Ref Target - Figure 4
99.9986%
100
99.9984%
50
99.9982%
0
99.9980%
KU040
Native Device
+ Essential Bits
+ Built-in
Correction
+ Partial Design
Level Mitigation
+ Full Design
Protection
WP462_04_020915
Figure 4: Increasing Device Availability through Additional SEU Mitigation
Built-In Detection and Correction
To enhance the SEU mitigation capabilities of UltraScale devices, the configuration logic includes a
built-in function that continuously scans the device configuration RAM to detect and correct
single- or multiple-bit upsets. Detection and correction of soft errors in UltraScale devices is
2X faster than prior generation FPGAs, allowing upsets to be corrected in just a few milliseconds
while the device continues operating in user mode.
Enhanced ECC is embedded in each configuration frame, enabling up to 8-bit error detection and
4-bit error correction per frame. Coupled with the physical frame interleaving mentioned
previously, this means a device can detect a multiple-bit upset up to 16 bits and correct up to 8 bits.
UltraScale devices are designed to ensure complete and robust coverage. In addition to the frame
ECC capabilities, a 32-bit CRC is calculated for the entire device configuration RAM, which can
reliably detect up to 31 randomized bit errors. The error detection capabilities of UltraScale devices
are exceptionally strong.
Improving the error detect and correct scan time by 2X minimizes system impact in the unlikely
event of an SEU event that results in a functional change to the design. SEU correction occurs
transparently, leveraging Xilinx's proven glitch-free partial reconfiguration technology, with no
impact to the user design other than the restoration of the intended functionality.
WP462 (v1.0) February 26, 2015
www.xilinx.com
6
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
The error mitigation clock that determines the scan time can be provided by the built-in clock
source or externally from the user design. The built-in mitigation logic also outputs a heartbeat
signal. This ensures the built-in function itself is functioning properly.
Soft Error Mitigation IP
To expand on the capabilities of the built-in detection and correction logic, Soft Error Mitigation
(SEM) IP is freely available to UltraScale FPGA customers. This small footprint IP core adds
advanced mitigation and verification capabilities to the solution, including:
•
Essential bits error classification and reporting
•
Fault injection for validating device and system response
•
Enhanced correction capabilities
Essential bits are defined as the set of configuration RAM bits in a given design that have any
possibility of creating a functional change to the design in the event of a change in state. In a
typical design that utilizes 70% or more of the FPGA resources, only 25% to 50% of configuration
RAM bits fall into the essential bits classification.
The Vivado® Design Suite allows for the creation of an essential bits mask file, which can be stored
in an off-chip flash memory. The SEM IP can use this mask file to compare detected configuration
RAM upsets against the known essential bits to classify and report the impact of the event. If no
essential bits are impacted, the user can choose to simply correct the configuration RAM bit and
continue system operation without further mitigation. This reduces the design-specific SEU FIT by
50% or more, compared to the intrinsic device SEU FIT.
The essential bits technology is intentionally conservative in its classification of configuration RAM
bits that might affect the design. Testing has shown that, in practice, even upsets to essential bits
have only a 10% to 30% probability of affecting design functionality.
Validation of system response to SEUs can be time consuming and costly. The UltraScale SEM IP
includes fault injection features to assist designers in determining system response to SEUs. Using
this knowledge, users can develop and verify fault mitigation algorithms at the system level. In
addition, essential bit masking technology can be combined with fault injection to target those
areas of the configuration RAM that could potentially affect the design functionality, thus further
reducing verification time.
As with previous generation devices, SEM IP remains in pre-production status until it has fully
passed accelerated radiation testing to ensure its robustness, at which time Xilinx provides test
data and analysis reports upon request.
WP462 (v1.0) February 26, 2015
www.xilinx.com
7
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Design Techniques
For applications requiring the very highest levels of design integrity, additional design-specific SEU
FIT reduction can be achieved with various design techniques to reduce or even eliminate changes
in device functionality from SEUs. These techniques generally involve building some form of partial
or total redundancy into the design to ensure that there is no single point of failure in critical
design modules.
When redundancy is not required, it is valuable to use Vivado Design Suite's hierarchical design
flow to restrict the placement of logic and routing as a means of separating critical design modules.
This methodology eliminates the possibility of an SEU event impacting multiple design modules
and allows designers to effectively analyze modules in isolation through focused fault injection on
only the applicable device area.
Analysis and Verification
Xilinx is the only vendor to openly publish SEU FIT/Mb data. This information is reported in UG116,
Device Reliability Report. Xilinx also assists third parties interested in validating the reported
results. In addition to this published information on SEU FIT/Mb for each device family and process
node, Xilinx provides tools and unique device capabilities to analyze and verify SEU mitigation
requirements specific to each customer's design.
Design Analysis
See Figure 5. For analysis in the early stages of a new design, Xilinx offers an SEU FIT Rate
Calculator to help designers assess device SEU FIT and plan for appropriate device and
system-level mitigation methods.
X-Ref Target - Figure 5
WP462_05_020915
Figure 5: SEU FIT Rate Calculator for Early-Stage Estimation
WP462 (v1.0) February 26, 2015
www.xilinx.com
8
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
When techniques such as partial triple module redundancy (pTMR) are used to protect critical
design modules, the Vivado Design Suite preserves the redundant modules while also performing
cross-clock domain timing analysis to assure the validity of the pTMR solution.
Designers can also assign an identifying tag to design modules, nets, primitives, or any other type
of design element. This identifying tag can be used in conjunction with essential bits technology to
enable different system responses, depending on how the design element impacted by a soft error
is classified. With this methodology, users can customize their SEU mitigation solutions to identify
and respond to SEUs in the most appropriate manner for a given application.
Design and Device Verification
After a design has been analyzed to determine the level of SEU mitigation required, and the
appropriate design techniques have been implemented to achieve the system requirements, a
thorough verification methodology must be implemented to validate the solution.
The UltraScale architecture SEU mitigation solution provides means of device and system validation
that are generally not possible in other technologies, such as with ASICs. UltraScale devices provide
an open platform for users to perform fault injection testing via the SEM IP or through custom
methods specific to a given design. This enables testing the response of the device and system to
soft errors, and offers the ability to validate the applied mitigation techniques without the expense
of costly accelerated radiation testing.
Xilinx actively assists customers with testing and validation of mitigation techniques implemented
in Xilinx devices. Through multiple product generations of experience, Xilinx has developed the
expertise to accurately predict real world SEU FIT/Mb for each new generation of products. This
effort begins with simulation modeling and silicon test vehicles prior to manufacturing the first
actual device.
After the first devices are received, verification of the expected device behavior is confirmed with
accelerated radiation testing. Xilinx takes this process one step further by collecting real world data
through real-time Rosetta testing (see WP286, Continuing Experiments of Atmospheric Neutron
Effects on Deep Submicron Integrated Circuits) of hundreds of devices at different latitudes,
longitudes, and altitudes around the world. This allows Xilinx to correlate the results from multiple
testing methodologies and provides customers with the highest confidence in the capabilities of
the products and the designs implemented in these products.
WP462 (v1.0) February 26, 2015
www.xilinx.com
9
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Conclusion
UltraScale devices offer industry-leading SEU resilience and mitigation through the most robust
and comprehensive solutions available to manage the effects of soft errors. Through proprietary
circuit and layout techniques, advanced mitigation IP, integrated tool flows, and the most complete
and open verification methodology in the industry, UltraScale FPGAs offer:
•
3X lower SEU FIT/Mb than prior generation devices
•
2X faster detection and correction of upsets
•
99.9% upset correction
•
Fault injection for design and test of device and system level mitigation
•
Built-in ECC on block RAM to protect on-chip user data
Xilinx is the leader in SEU mitigation and is the only SRAM FPGA vendor repeatedly shown to be
capable of supporting applications requiring the highest reliability, availability, and functional
safety standards. UltraScale devices are the next step in Xilinx's continuing efforts to offer the most
robust and comprehensive solutions available.
Related Reading
1. Wikipedia: Single Event Upset. Retrieved January 6th, 2014 from:
http://en.wikipedia.org/wiki/Single_event_upset
2. P. Rech, C. Aguiar, R. Ferreira, M. Silvestri, A. Griffoni, C. Frost, and L. Carro: Neutron-Induced Soft
Errors in Graphic Processing Units. Retrieved [date unavailable] from:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6353714
3. Michael Santarini: Cosmic Radiation Comes to ASIC and SOC Design. Retrieved May 12, 2005
from:
http://www.edn.com/design/integrated-circuit-design/4324957/Cosmic-radiation-comes-to-ASIC-and-SOC-design
4. WP402, Considerations Surrounding Single Event Effects in FPGAs, ASICs, and Processors (Xilinx
White Paper)
WP462 (v1.0) February 26, 2015
www.xilinx.com
10
UltraScale Devices Maximize Design Integrity with Industry-Leading SEU Resilience and Mitigation
Revision History
The following table shows the revision history for this document:
Date
Version
02/26/2015
1.0
Description of Revisions
Initial Xilinx release.
Disclaimer
The information disclosed to you hereunder (the “Materials”) is provided solely for the selection and use of Xilinx
products. To the maximum extent permitted by applicable law: (1) Materials are made available "AS IS" and with all faults,
Xilinx hereby DISCLAIMS ALL WARRANTIES AND CONDITIONS, EXPRESS, IMPLIED, OR STATUTORY, INCLUDING BUT NOT
LIMITED TO WARRANTIES OF MERCHANTABILITY, NON-INFRINGEMENT, OR FITNESS FOR ANY PARTICULAR PURPOSE;
and (2) Xilinx shall not be liable (whether in contract or tort, including negligence, or under any other theory of liability)
for any loss or damage of any kind or nature related to, arising under, or in connection with, the Materials (including your
use of the Materials), including for any direct, indirect, special, incidental, or consequential loss or damage (including loss
of data, profits, goodwill, or any type of loss or damage suffered as a result of any action brought by a third party) even
if such damage or loss was reasonably foreseeable or Xilinx had been advised of the possibility of the same. Xilinx
assumes no obligation to correct any errors contained in the Materials or to notify you of updates to the Materials or to
product specifications. You may not reproduce, modify, distribute, or publicly display the Materials without prior written
consent. Certain products are subject to the terms and conditions of Xilinx’s limited warranty, please refer to Xilinx’s
Terms of Sale which can be viewed at http://www.xilinx.com/legal.htm#tos; IP cores may be subject to warranty and
support terms contained in a license issued to you by Xilinx. Xilinx products are not designed or intended to be fail-safe
or for use in any application requiring fail-safe performance; you assume sole risk and liability for use of Xilinx products
in such critical applications, please refer to Xilinx’s Terms of Sale which can be viewed at http://www.xilinx.com/
legal.htm#tos.
Automotive Applications Disclaimer
XILINX PRODUCTS ARE NOT DESIGNED OR INTENDED TO BE FAIL-SAFE, OR FOR USE IN ANY APPLICATION REQUIRING
FAIL-SAFE PERFORMANCE, SUCH AS APPLICATIONS RELATED TO: (I) THE DEPLOYMENT OF AIRBAGS, (II) CONTROL OF A
VEHICLE, UNLESS THERE IS A FAIL-SAFE OR REDUNDANCY FEATURE (WHICH DOES NOT INCLUDE USE OF SOFTWARE IN
THE XILINX DEVICE TO IMPLEMENT THE REDUNDANCY) AND A WARNING SIGNAL UPON FAILURE TO THE OPERATOR, OR
(III) USES THAT COULD LEAD TO DEATH OR PERSONAL INJURY. CUSTOMER ASSUMES THE SOLE RISK AND LIABILITY OF
ANY USE OF XILINX PRODUCTS IN SUCH APPLICATIONS.
WP462 (v1.0) February 26, 2015
www.xilinx.com
11
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising