7 Series FPGAs Configurable Logic Block User Guide (UG474)

7 Series FPGAs Configurable Logic Block User Guide (UG474)
7 Series FPGAs
Configurable Logic Block
User Guide
UG474 (v1.7) November 17, 2014
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, NONINFRINGEMENT, 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 failsafe 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.
© Copyright 2011–2014 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.
Revision History
The following table shows the revision history for this document.
Date
Version
Revision
03/01/2011
1.0
Xilinx Initial release.
03/28/2011
1.1
Added devices XC7K355T, XC7K420T, and XC7K480T to Table 1-2. Portions of the text
have been revised for clarity.
09/30/2011
1.2
Added last sentence under 7 Series CLB Features. Added Table 1-2 and Table 1-3.
Updated CLB features in Table 1-2. Added first sentence under CLB Arrangement,
added ASMBL Architecture section, and CLB Slices heading. Added last paragraph
under Carry Logic. Added lasts sentence under Using Carry Logic. Added second
sentence under Slice Multiplexer Timing Parameters. Modified Table 5-2, Table 5-4, and
Table 5-5 for clarity. Added Devices Using Stacked Silicon Interconnect (SSI) Technology
section.
01/30/2012
1.3
Revised Table 1-1. Added fifth paragraph under Distributed RAM (Available in SLICEM
Only). Clarified last paragraph under Global Controls GSR and GTS.
11/05/2012
1.4
Changed “uniformity” to “optimized” in last bullet under 7 Series CLB Features.
Changed “unified” to “scalable” in first sentence under Device Resources. Deleted
7A350T device from Table 1-1. Deleted 7V1500T and 7VH290T devices from Table 1-3.
Added reference to 7 Series FPGA Libraries Guide to Distributed RAM (Available in
SLICEM Only), Shift Registers (Available in SLICEM Only), and Flip-Flop Primitives.
Changed “TCEO” to “TCECK” in Figure 5-2 and first bullet under General Timing
Characteristics.
7 Series FPGAs CLB User Guide
www.xilinx.com
UG474 (v1.7) November 17, 2014
Date
Version
Revision
08/6/2013
1.5
Added Artix®-7 devices. Updated references to implementation tools.
08/11/2014
1.6
Revised footnotes in Table 1-1 through Table 1-3. Revised polarity from independent to
programmable in Control Signals, page 22. Added Primitive column to Table 2-3 and
removed footnotes. Renamed or made minor revisions to Figure 2-6 through Figure 2-14.
Revised sections Clock – WCLK, page 49, Clock – CLK, page 50, and Clock - C, page 51.
11/17/2014
1.7
Updated Table 1-1 for new Artix 7A15T device.
UG474 (v1.7) November 17, 2014
www.xilinx.com
7 Series FPGAs CLB User Guide
7 Series FPGAs CLB User Guide
www.xilinx.com
UG474 (v1.7) November 17, 2014
Table of Contents
Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Preface: About This Guide
Guide Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Additional Support Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Chapter 1: Overview
CLB Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
7 Series CLB Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Device Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Recommended Design Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Pinout Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Chapter 2: Functional Details
CLB Arrangement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Slice Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Look-Up Table (LUT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Storage Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Distributed RAM (Available in SLICEM Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Shift Registers (Available in SLICEM Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Multiplexers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Carry Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
18
21
21
23
34
39
43
Chapter 3: Design Entry
Design Checklist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Using the CLB Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Primitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Chapter 4: Applications
Distributed RAM Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Shift Register Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Carry Logic Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Chapter 5: Timing
CLB General Slice Timing Model and Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
CLB Slice Multiplexer Timing Model and Parameters . . . . . . . . . . . . . . . . . . . . . . . . 60
CLB Slice Carry-Chain Timing Model and Parameters . . . . . . . . . . . . . . . . . . . . . . . 61
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
5
CLB Slice Distributed RAM Timing Model and Parameters
(Available in SLICEM Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
CLB Slice SRL Shift Register Timing Model and Parameters
(Available in SLICEM Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Chapter 6: Advanced Topics
Using the Latch Function as Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Interconnect Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Devices Using Stacked Silicon Interconnect (SSI) Technology . . . . . . . . . . . . . . . 73
6
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Preface
About This Guide
Xilinx® 7 series FPGAs include three FPGA families that are all designed for lowest power
to enable a common design to scale across families for optimal power, performance, and
cost. The Artix®-7 family is optimized for lowest cost and absolute power for the highest
volume applications. The Virtex®-7 family is optimized for highest system performance
and capacity. The Kintex®-7 family is an innovative class of FPGAs optimized for the best
price-performance. This guide serves as a technical reference describing the 7 series FPGAs
configurable logic blocks (CLBs).
Usually, logic synthesis assigns the CLB resources without system designer intervention. It
can be advantageous for the designer to understand certain CLB details, including the
varying capabilities of the look-up tables (LUTs), the physical direction of the carry
propagation, the number and distribution of the available flip-flops, and the availability of
the very efficient shift registers. This guide describes these and other features of the CLB in
detail.
This 7 Series FPGAs Configurable Logic Block User Guide, part of an overall set of
documentation on the 7 series FPGAs, is available on the Xilinx 7 Series documentation
website.
Guide Contents
This manual contains these chapters:
•
Chapter 1, Overview, provides basic information needed for the majority of users,
including:
•
CLB Overview is targeted at the new user.
•
7 Series CLB Features discusses what is new compared with the Spartan®-6 and
Virtex®-6 FPGA families for the experienced user and provides design migration
considerations.
•
Device Resources indicates the number of resources per device, and unity
between different 7 series families.
•
Recommended Design Flow provides the basics of using CLB resources and lists
key aspects to consider.
•
Pinout Planning discusses aspects of CLBs that might affect pin placement for a
design.
•
Chapter 2, Functional Details, lists architectural specifics for each CLB feature.
•
Chapter 3, Design Entry, provides design entry guidelines and primitives for
instantiation.
•
Chapter 4, Applications, provides examples that use the CLB resources in larger
applications.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
7
Preface: About This Guide
•
Chapter 5, Timing, contains timing models and defines CLB timing specifications
from the respective 7 series FPGA data sheet.
•
Chapter 6, Advanced Topics, discusses advanced features of the 7 series CLB.
Additional Support Resources
To find additional documentation, see the Xilinx website at:
www.xilinx.com/support/documentation/index
To search the Answer Database of silicon, software, and IP questions and answers, or to
create a technical support WebCase, see the Xilinx website at:
www.xilinx.com/support
8
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 1
Overview
CLB Overview
The 7 series configurable logic block (CLB) provides advanced, high-performance FPGA
logic:
•
Real 6-input look-up table (LUT) technology
•
Dual LUT5 (5-input LUT) option
•
Distributed Memory and Shift Register Logic capability
•
Dedicated high-speed carry logic for arithmetic functions
•
Wide multiplexers for efficient utilization
CLBs are the main logic resources for implementing sequential as well as combinatorial
circuits. Each CLB element is connected to a switch matrix for access to the general routing
matrix (shown in Figure 1-1). A CLB element contains a pair of slices.
X-Ref Target - Figure 1-1
COUT
COUT
CLB
Slice(1)
Switch
Matrix
Slice(0)
CIN
CIN
UG474_c1_01_071910
Figure 1-1:
Arrangement of Slices within the CLB
The LUTs in 7 series FPGAs can be configured as either a 6-input LUT with one output, or
as two 5-input LUTs with separate outputs but common addresses or logic inputs. Each
5-input LUT output can optionally be registered in a flip-flop. Four such 6-input LUTs and
their eight flip-flops as well as multiplexers and arithmetic carry logic form a slice, and two
slices form a CLB. Four flip-flops per slice (one per LUT) can optionally be configured as
latches. In that case, the remaining four flip-flops in that slice must remain unused.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
9
Chapter 1: Overview
Approximately two-thirds of the slices are SLICEL logic slices and the rest are SLICEM,
which can also use their LUTs as distributed 64-bit RAM or as 32-bit shift registers (SRL32)
or as two SRL16s. Modern synthesis tools take advantage of these highly efficient logic,
arithmetic, and memory features. Expert designers can also instantiate them.
7 Series CLB Features
The 7 series CLB is identical to that in the Virtex®-6 FPGA family. The CLB is very similar
to that of the Spartan®-6 FPGA family with these differences:
•
•
Columnar architecture
•
Scales easily to higher densities
•
More routing between CLBs
SLICEL and SLICEM only (no Spartan-6 FPGA SLICEX)
•
All slices support carry logic
•
More optimized
The common features in the CLB structure simplify design migration from the Spartan-6
and Virtex-6 families to the 7 series devices. The unique floorplan means that location
constraints should be removed before implementing designs originally targeted to earlier
FPGAs. The interconnect routing resources are increased in size, quantity, and flexibility
relative to the Virtex-6 FPGA family, improving the quality of automatic place and route
results.
Device Resources
The CLB resources are scalable across all the 7 series families, providing a common
architecture that improves efficiency, IP implementation, and design migration. The
number of CLBs and the ratio between CLBs and other device resources differentiates the
7 series families. Migration between the 7 series families does not require any design
changes for the CLBs.
Device capacity is often measured in terms of logic cells, which are the logical equivalent of
a classic four-input LUT and a flip-flop. The 7 series FPGA CLB six-input LUT, abundant
flip-flops and latches, carry logic, and the ability to create distributed RAM or shift
registers in the SLICEM, increase the effective capacity. The ratio between the number of
logic cells and 6-input LUTs is 1.6:1.
7 Series FPGA CLB Resources
Table 1-1 through Table 1-3 show the available CLB resources for the Artix®-7, Kintex®-7,
and Virtex®-7 FPGAs. Refer to DS180, 7 Series FPGAs Overview for the most up-to-date
information.
Table 1-1:
10
Artix-7 FPGA CLB Resources
Device
Slices(1)
SLICEL
SLICEM
6-input
LUTs
Distributed RAM
(Kb)
Shift
Register
(Kb)
Flip-Flops
7A15T
2,600(2)
1,800
800
10,400
200
100
20,800
7A35T
5,200(2)
3,600
1,600
20,800
400
200
41,600
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Device Resources
Table 1-1:
Artix-7 FPGA CLB Resources (Cont’d)
Device
Slices(1)
SLICEL
SLICEM
6-input
LUTs
Distributed RAM
(Kb)
Shift
Register
(Kb)
Flip-Flops
7A50T
8,150
5,750
2,400
32,600
600
300
65,200
7A75T
11,800(2)
8,232
3,568
47,200
892
446
94,400
7A100T
15,850
11,100
4,750
63,400
1,188
594
126,800
7A200T
33,650
22,100
11,550
134,600
2,888
1,444
269,200
Notes:
1. Each 7 series FPGA slice contains four LUTs and eight flip-flops; only SLICEMs can use their LUTs as distributed RAM or SRLs.
2. Number of slices corresponding to the number of LUTs and flip-flops supported in the device.
Table 1-2:
Kintex-7 FPGA CLB Resources
Device
Slices(1)
SLICEL
SLICEM
6-input
LUTs
Distributed RAM
(Kb)
Shift
Register
(Kb)
Flip-Flops
7K70T
10,250
6,900
3,350
41,000
838
419
82,000
7K160T
25,350
16,600
8,750
101,400
2,188
1,094
202,800
7K325T
50,950
34,950
16,000
203,800
4,000
2,000
407,600
7K355T
55,650
35,300
20,350
222,600
5,088
2,544
445,200
7K410T
63,550
40,900
22,650
254,200
5,663
2,831
508,400
41,400
23,750
260,600
5,938
2,969
521,200
47,500
27,150
298,600
6,788
3,394
597,200
7K420T
7K480T
(2)
65,150
74,650
Notes:
1. Each 7 series FPGA slice contains four LUTs and eight flip-flops; only SLICEMs can use their LUTs as distributed RAM or SRLs.
2. Number of slices corresponding to the number of LUTs and flip-flops supported in the device.
Table 1-3:
Virtex-7 FPGA CLB Resources
Device
Slices(1)
SLICEL
SLICEM
6-input
LUTs
Distributed RAM
(Kb)
Shift
Register
(Kb)
Flip-Flops
7V585T
91,050
63,300
27,750
364,200
6,938
3,469
728,400
7V2000T
305,400
219,200
86,200
1,221,600
21,550
10,775
2,443,200
7VX330T
51,000
33,450
17,550
204,000
4,388
2,194
408,000
7VX415T
64,400
38,300
26,100
257,600
6,525
3,263
515,200
7VX485T
75,900
43,200
32,700
303,600
8,175
4,088
607,200
7VX550T
86,600(2)
51,700
34,900
346,400
8,725
4,363
692,800
7VX690T
108,300
64,750
43,550
433,200
10,888
5,444
866,400
7VX980T
153,000
97,650
55,350
612,000
13,838
6,919
1,224,000
7VX1140T
178,000
107,200
70,800
712,000
17,700
8,850
1,424,000
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
11
Chapter 1: Overview
Table 1-3:
Virtex-7 FPGA CLB Resources (Cont’d)
Device
Slices(1)
SLICEL
SLICEM
6-input
LUTs
Distributed RAM
(Kb)
Shift
Register
(Kb)
Flip-Flops
7VH580T
90,700
55,300
35,400
362,800
8,850
4,425
725,600
7VH870T
136,900
83,750
53,150
547,600
13,275
6,638
1,095,200
Notes:
1. Each 7 series FPGA slice contains four LUTs and eight flip-flops; only SLICEMs can use their LUTs as distributed RAM or SRLs.
2. Number of slices corresponding to the number of LUTs and flip-flops supported in the device.
Recommended Design Flow
CLB resources are inferred for generic design logic and do not require instantiation. Good
HDL design is sufficient. A few items to note:
•
CLB flip-flops have either a set or a reset. The designer must not use both set and
reset.
•
Flip-flops are abundant. Pipelining should be considered to improve performance.
•
Control inputs are shared across a slice or CLB. The number of unique control inputs
required for a design should be minimized. Control inputs include clock, clock enable,
set/reset, and write enable.
•
A 6-input LUT can be used as a 32-bit shift register for efficient implementation.
•
A 6-input LUT can be used as a 64 x 1 memory for small storage requirements.
•
Dedicated carry logic implements arithmetic functions effectively.
These steps indicate the recommended design flow:
1.
Implement the design using preferred methodologies (HDL, IP, etc.).
2.
Evaluate utilization reports to determine resources used.
Check to make sure arithmetic logic, distributed RAM, and SRL are used, when
helpful.
3.
4.
Consider flip-flop usage.
a.
Pipeline for performance
b.
Use dedicated flip-flops at the outputs of dedicated resources (block RAM, DSP)
c.
Allow shift registers to use SRL (avoid set/resets)
Minimize the use of set/resets.
Pinout Planning
Although the use of most resources affects the resulting device pinout, CLB usage has little
effect on pinouts because they are distributed throughout the device. The ASMBL™
architecture provides maximum flexibility with CLBs on both sides of most I/Os.
The best approach is to let the tools choose the I/O locations based on the FPGA
requirements. Results can be adjusted if necessary for board layout considerations. The
timing constraints should be set so that the tools can choose optimal placement for the
design requirements.
12
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Pinout Planning
Carry logic cascades vertically up a column, so wide arithmetic buses might drive a
vertical orientation to other logic, including I/O.
While most 7 series devices are available in flip-chip packages, taking full advantage of the
distributed I/O in the ASMBL architecture, the smaller devices are available in wire-bond
packages at a lower cost. In these packages, some pins are naturally closer to the I/Os and
special resources than others, so pin placement should be done after the internal logic is
defined.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
13
Chapter 1: Overview
14
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 2
Functional Details
This chapter provides a detailed view of the 7 series FPGAs CLB architecture. These details
can be useful for design optimization and verification, but are not necessary for initiating
a design. This chapter includes:
•
CLB Arrangement
Overview of slice locations and features within the CLB
•
Slice Description
Complete details of SLICEM and SLICEL
•
Look-Up Table (LUT)
Description of the logical function generators
•
Storage Elements
Description and controls for the latches and flip-flops
•
Distributed RAM (Available in SLICEM Only)
SLICEM ability to use LUTs as writable memory
•
Shift Registers (Available in SLICEM Only)
SLICEM ability to use LUTs as shift registers
•
Multiplexers
Dedicated gates for combining LUTs into wide functions
•
Carry Logic
Dedicated gates and cascading to implement efficient arithmetic functions
CLB Arrangement
The CLBs are arranged in columns in the 7 series FPGAs. The 7 series is the fourth
generation to be based on the unique columnar approach provided by the ASMBL™
architecture.
ASMBL Architecture
Xilinx created the Advanced Silicon Modular Block (ASMBL) architecture to enable FPGA
platforms with varying feature mixes optimized for different application domains.
Through this innovation Xilinx offers a greater selection of devices, enabling customers to
select the FPGA with the right mix of features and capabilities for their specific design.
Figure 2-1 provides a high-level description of the different types of column-based
resources.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
15
Chapter 2: Functional Details
X-Ref Target - Figure 2-1
Feature Options
Column
Based
ASMBL
Architecture
Logic (SLICEL)
Global Clock
Logic (SLICEM)
High-performance I/O
DSP
High-range I/O
Memory
Integrated IP
Clock Management Tile
Mixed Signal
Transceivers
Domain A
Domain B
Domain C
Applications
Applications
Applications
UG474_c2_24_071014
Figure 2-1:
ASMBL Architecture
The ASMBL architecture breaks through traditional design barriers by:
•
Eliminating geometric layout constraints such as dependencies between I/O count
and array size.
•
Enhancing on-chip power and ground distribution by allowing power and ground to
be placed anywhere on the chip.
•
Allowing disparate integrated IP blocks to be scaled independent of each other and
surrounding resources.
SSI Technology
The 7 series FPGAs extend integration even higher by using the unique stacked silicon
interconnect (SSI) technology. SSI technology enables multiple super logic regions (SLRs)
to be combined on a passive interposer layer, to create a single FPGA with more than ten
thousand inter-SLR connections. See Chapter 6, Advanced Topics for additional
information.
CLB Slices
A CLB element contains a pair of slices, and each slice is composed of four 6-input LUTs
and eight storage elements.
•
SLICE(0) – slice at the bottom of the CLB and in the left column
•
SLICE(1) – slice at the top of the CLB and in the right column
These two slices do not have direct connections to each other, and each slice is organized as
a column. Each slice in a column has an independent carry chain.
16
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Arrangement
The Xilinx tools designate slices with these definitions:
•
An “X” followed by a number identifies the position of each slice in a pair as well as
the column position of the slice. The “X” number counts slices starting from the
bottom in sequence 0, 1 (the first CLB column); 2, 3 (the second CLB column); etc.
•
A “Y” followed by a number identifies a row of slices. The number remains the same
within a CLB, but counts up in sequence from one CLB row to the next CLB row,
starting from the bottom.
Figure 2-2 shows four CLBs located in the bottom-left corner of the die.
X-Ref Target - Figure 2-2
COUT
COUT
CLB
COUT
COUT
CLB
Slice1
X1Y1
Slice0
X0Y1
Slice1
X3Y1
Slice0
X2Y1
CIN
CIN
CIN
CIN
COUT
COUT
COUT
COUT
CLB
CLB
Slice1
X1Y0
Slice0
X0Y0
Slice1
X3Y0
Slice0
X2Y0
UG474_c2_01_092210
Figure 2-2:
Row and Column Relationship between CLBs and Slices
CLB/Slice Configurations
Table 2-1 summarizes the logic resources in one CLB. Each SLICEM LUT can be configured
as a look-up table, distributed RAM, or a shift register.
Table 2-1:
Logic Resources in One CLB
Slices
LUTs
Flip-Flops
Arithmetic and
Carry Chains
Distributed RAM(1)
Shift Registers(1)
2
8
16
2
256 bits
128 bits
Notes:
1. SLICEM only, SLICEL does not have distributed RAM or shift registers.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
17
Chapter 2: Functional Details
Slice Description
Every slice contains:
•
Four logic-function generators (or look-up tables)
•
Eight storage elements
•
Wide-function multiplexers
•
Carry logic
These elements are used by all slices to provide logic, arithmetic, and ROM functions. In
addition, some slices support two additional functions: storing data using distributed
RAM and shifting data with 32-bit registers. Slices that support these additional functions
are called SLICEM; others are called SLICEL. SLICEM (shown in Figure 2-3) represents a
superset of elements and connections found in all slices. SLICEL is shown in Figure 2-4.
Each CLB can contain two SLICEL or a SLICEL and a SLICEM.
18
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Slice Description
X-Ref Target - Figure 2-3
D
CE
CK
COUT
SRHI
SRLO
INIT1 Q
INIT0
SR
Reset Type
Sync/Async
FF/LAT
DX
D6:1
DMUX
DI2
A6:A1
W6:W1
CK
D
D
O6
O5
DX
D
CE
CK
DI1
WEN MC31
D
DI
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
FF/LAT
INIT1 Q
INIT0
SRHI
SRLO
SR
CX
C6:1
CMUX
DI2
A6:A1
W6:W1
CK
C
C
O6
O5
CX
D
CE
CK
DI1
WEN MC31
D
CI
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
FF/LAT
INIT1 Q
INIT0
SRHI
SRLO
BMUX
DI2
A6:A1
W6:W1
CK
B
B
O6
O5
BX
DI1
WEN MC31
D
BI
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
D
CE
CK
FF/LAT
INIT1 Q
INIT0
SRHI
SRLO
BQ
SR
AX
A6:1
CQ
SR
BX
B6:1
DQ
AMUX
DI2
A6:A1
W6:W1
CK
AX
DI1
WEN MC31
AI
A
A
O6
O5
D
CE
CK
FF/LAT
INIT1 Q
INIT0
SRHI
SRLO
AQ
SR
0/1
SR
CE
CLK
CK
WE
WEN
CIN
UG474_c2_02_110510
Figure 2-3:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Diagram of SLICEM
www.xilinx.com
Send Feedback
19
Chapter 2: Functional Details
X-Ref Target - Figure 2-4
D
COUT
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
Reset Type
Sync/Async
FF/LAT
DX
D6:1
DMUX
A6:A1
D
D
O6
O5
DX
D
CE
CK
D
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
FF/LAT
INIT1
Q
INIT0
SRHI
SRLO
SR
CX
C6:1
CMUX
A6:A1
C
C
O6
O5
CX
D
CE
CK
D
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
FF/LAT
INIT1
Q
INIT0
SRHI
SRLO
BMUX
A6:A1
B
B
O6
O5
BX
D
CE
CK
SRHI
SRLO
INIT1 Q
INIT0
SR
D
CE
CK
FF/LAT
Q
INIT1
INIT0
SRHI
SRLO
SR
AX
A6:1
CQ
SR
BX
B6:1
DQ
BQ
AMUX
A6:A1
A
A
O6
O5
AX
D
CE
CK
FF/LAT
Q
INIT1
INIT0
SRHI
SRLO
AQ
SR
0/1
SR
CE
CLK
CIN
UG474_c2_03_101210
Figure 2-4: Diagram of SLICEL
20
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Look-Up Table (LUT)
Look-Up Table (LUT)
The function generators in 7 series FPGAs are implemented as six-input look-up tables
(LUTs). There are six independent inputs (A inputs - A1 to A6) and two independent
outputs (O5 and O6) for each of the four function generators in a slice (A, B, C, and D). The
function generators can implement:
•
Any arbitrarily defined six-input Boolean function
•
Two arbitrarily defined five-input Boolean functions, as long as these two functions
share common inputs
•
Two arbitrarily defined Boolean functions of 3 and 2 inputs or less
A six-input function uses:
•
A1-A6 inputs
•
O6 output
Two five-input or less functions use:
•
A1–A5 inputs
•
A6 driven High
•
O5 and O6 outputs
The propagation delay through a LUT is independent of the function implemented.
Signals from the function generators can:
•
Exit the slice (through A, B, C, D output for O6 or AMUX, BMUX, CMUX, DMUX
output for O5)
•
Enter the XOR dedicated gate from an O6 output
•
Enter the carry-logic chain from an O5 output
•
Enter the select line of the carry-logic multiplexer from O6 output
•
Feed the D input of the storage element
•
Go to F7AMUX/F7BMUX wide multiplexers from O6 output
In addition to the basic LUTs, slices contain three multiplexers (F7AMUX, F7BMUX, and
F8MUX). These multiplexers are used to combine up to four function generators to provide
any function of seven or eight inputs in a slice.
•
F7AMUX: Used to generate seven input functions from LUTs A and B
•
F7BMUX: Used to generate seven input functions from LUTs C and D
•
F8MUX: Used to combine all LUTs to generate eight input functions.
Functions with more than eight inputs can be implemented using multiple slices. There are
no direct connections between slices to form function generators greater than eight inputs
within a CLB.
Storage Elements
There are eight storage elements per slice. Four can be configured as either edge-triggered
D-type flip-flops or level-sensitive latches. The D input can be driven directly by a LUT
output via AFFMUX, BFFMUX, CFFMUX, or DFFMUX, or by the BYPASS slice inputs
bypassing the function generators via AX, BX, CX, or DX input. When configured as a
latch, the latch is transparent when the CLK is Low.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
21
Chapter 2: Functional Details
There are four additional storage elements that can only be configured as edge-triggered
D-type flip-flops. The D input can be driven by the O5 output of the LUT or the BYPASS
slice inputs via AX, BX, CX, or DX input. When the original four storage elements are
configured as latches, these four additional storage elements cannot be used.
Figure 2-5 shows both the register only and the register/latch configuration in a slice.
X-Ref Target - Figure 2-5
DFF
LUT D O5 Output
D
CE
CK
DX
INIT1
Q
INIT0
SRHIGH
SRLOW
DQ
D
CE
CK
CX
D
CE
CK
DX
SR
CFF
LUT C O5 Output
DFF/LATCH
LUT D Output
CQ
D
CE
CK
CX
SR
Reset Type
SR
FF
LATCH
Q
INIT1
INIT0
SRHIGH
SRLOW
SR
BFF
BX
D
CE
CK
CE
CLK
INIT1
Q
INIT0
SRHIGH
SRLOW
Sync
LUT B Output
Async
BQ
BFF/LATCH
BX
D
CE
CK
AX
INIT1
Q
INIT0
SRHIGH
SRLOW
D
CE
CK
CE
SR
CLK
AFF
LUT A O5 Output
CQ
Reset Type
SR
Sync
LUT B O5 Output
DQ
CFF/LATCH
LUT C Output
INIT1
Q
INIT0
SRHIGH
SRLOW
FF
LATCH
Q
INIT1
INIT0
SRHIGH
SRLOW
SR
Async
FF
LATCH
Q
INIT1
INIT0
SRHIGH
SRLOW
SR
BQ
AFF/LATCH
LUT A Output
AQ
D
CE
CK
AX
SR
FF
LATCH
Q
INIT1
INIT0
SRHIGH
SRLOW
SR
AQ
UG474_c2_04_101210
Figure 2-5:
Two Versions of Configuration in a Slice: 4 Registers Only and 4 Register/Latch
Control Signals
The control signals clock (CLK), clock enable (CE), and set/reset (SR) are common to all
storage elements in one slice. When one flip-flop in a slice has SR or CE enabled, the other
flip-flops used in the slice also have SR or CE enabled by the common signal. Only the CLK
signal has programmable polarity. Any inverter placed on the clock signal is automatically
absorbed. The CE and SR signals are active-High.
These initialization options are available for storage elements:
22
•
SRLOW: Synchronous or asynchronous Reset when CLB SR signal is asserted
•
SRHIGH: Synchronous or asynchronous Set when CLB SR signal is asserted
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
•
INIT0: Asynchronous Reset on power-up or global Set/Reset (see Global Controls
GSR and GTS, page 72)
•
INIT1: Asynchronous Set on power-up or global Set/Reset
The SR signal forces the storage element into the state specified by the SRHIGH or SRLOW
attribute. SRHIGH forces a logic High at the storage element output when SR is asserted,
while SRLOW forces a logic Low at the storage element output (see Table 2-2).
Table 2-2:
Truth Table when using SRLOW and SRHIGH
SR
SRVAL
Function
0
SRLOW (default)
No Logic Change
1
SRLOW (default)
0
0
SRHIGH
No Logic Change
1
SRHIGH
1
SRHIGH and SRLOW can be set individually for each storage element in a slice. The choice
of synchronous (SYNC) or asynchronous (ASYNC) set/reset (SRTYPE) cannot be set
individually for each storage element in a slice.
The initial state after configuration or global initial state is defined by separate INIT0 and
INIT1 attributes. By default, setting the SRLOW attribute sets INIT0, and setting the
SRHIGH attribute sets INIT1. 7 series devices can set INIT0 and INIT1 independent of
SRHIGH and SRLOW.
The configuration options for the set and reset functionality of a register or the four storage
elements capable of functioning as a latch are:
•
No set or reset
•
Synchronous set
•
Synchronous reset
•
Asynchronous set (preset)
•
Asynchronous reset (clear)
Distributed RAM (Available in SLICEM Only)
The function generators (LUTs) in SLICEMs can be implemented as a synchronous RAM
resource called a distributed RAM element. Multiple LUTs in a SLICEM can be combined
in various ways to store larger amount of data. RAM elements are configurable within a
SLICEM to implement these configurations:
•
Single-Port 32 x 1-bit RAM
•
Dual-Port 32 x 1-bit RAM
•
Quad-Port 32 x 2-bit RAM
•
Simple Dual-Port 32 x 6-bit RAM
•
Single-Port 64 x 1-bit RAM
•
Dual-Port 64 x 1-bit RAM
•
Quad-Port 64 x 1-bit RAM
•
Simple Dual-Port 64 x 3-bit RAM
•
Single-Port 128 x 1-bit RAM
•
Dual-Port 128 x 1-bit RAM
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
23
Chapter 2: Functional Details
•
Single-Port 256 x 1-bit RAM
Distributed RAM modules are synchronous (write) resources. A synchronous read can be
implemented with a flip-flop in the same slice. By using this flip-flop, the distributed RAM
performance is improved by decreasing the delay into the clock-to-out value of the
flip-flop. However, an additional clock latency is added. The distributed elements share
the same clock input. For a write operation, the Write Enable (WE) input, driven by either
the CE or WE pin of a SLICEM, must be set High.
Table 2-3 shows the number of LUTs (four per slice) occupied by each distributed RAM
configuration. See UG953, Vivado Design Suite 7 Series FPGA and Zynq-7000 All
Programmable SoC Libraries Guide for details of available distributed RAM primitives.
Table 2-3:
Distributed RAM Configuration
RAM
Description
Primitive
Number of LUTs
32 x 1S
Single port
RAM32X1S
1
32 x 1D
Dual port
RAM32X1D
2
32 x 2Q
Quad port
RAM32M
4
32 x 6SDP
Simple dual port
RAM32M
4
64 x 1S
Single port
RAM64X1S
1
64 x 1D
Dual port
RAM64X1D
2
64 x 1Q
Quad port
RAM64M
4
64 x 3SDP
Simple dual port
RAM64M
4
128 x 1S
Single port
RAM128X1S
2
128 x 1D
Dual port
RAM128X1D
4
256 x 1S
Single port
RAM256X1S
4
Distributed RAM configurations include:
•
Single port
•
Common address port for synchronous writes and asynchronous reads
-
•
Dual port
•
One port for synchronous writes and asynchronous reads
-
•
Send Feedback
One function generator is connected with the shared read and write port
address
One port for asynchronous reads
-
24
Read and write addresses share the same address bus
Second function generator has the A inputs connected to a second read-only
port address, and the WA inputs are shared with the first read/write port
address
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
•
•
Simple dual port
•
One port for synchronous writes (no data out/read port from the write port)
•
One port for asynchronous reads
Quad port
•
One port for synchronous writes and asynchronous reads
•
Three ports for asynchronous reads
As shown in Figure 2-3, the common write port W6:W1 (WA[6:1] in the following figures)
is always physically driven by the inputs to the D LUT using D[6:1]. The read ports are
independent for each of the four LUTs. Therefore the D LUT is always effectively
single port, even if DPRAM64 is selected as the LUT configuration. The other three LUTs
are always effectively dual port, although SPRAM32 can be selected when the read and
write addresses are connected together.
Figure 2-6 through Figure 2-14 illustrate various example distributed RAM configurations
occupying one SLICEM.
When using x2 configurations (as in 32 X 2 Quad Port in Figure 2-6), A6 and WA6 are
driven High by the software to keep O5 and O6 independent.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
25
Chapter 2: Functional Details
X-Ref Target - Figure 2-6
RAM32M
DID[1]
DID[0]
ADDRD[4:0]
(DI)
(AX/BX/CX/DX)
D[5:1] 5
5
(CLK)
(WE)
WCLK
WED
DPRAM32
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
DOD[0]
O5
DOD[1]
DPRAM32
ADDRC[4:0]
C[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
DOC[0]
O5
DOC[1]
DPRAM32
ADDRB[4:0]
B[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
DOB[0]
O5
DOB[1]
DPRAM32
ADDRA[4:0]
A[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
DOA[0]
O5
DOA[1]
UG474_c2_06_070914
Figure 2-6:
26
Send Feedback
32 X 2 Quad Port Distributed RAM (RAM32M)
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
X-Ref Target - Figure 2-7
RAM32M
DPRAM32
unused
unused
WADDR[5:1]
WADDR[6] = 1
D[5:1] 5
5
(CLK)
(WE)
WCLK
WED
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
DPRAM32
DATA[1]
DATA[2]
RADDR[5:1]
RADDR[6] = 1
C[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[2]
O5
O[1]
DPRAM32
DATA[3]
DATA[4]
B[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[4]
O5
O[3]
DPRAM32
DATA[5]
DATA[6]
A[5:1] 5
5
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[6]
O5
O[5]
UG474_c2_06_070914
Figure 2-7:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
32 X 6 Simple Dual Port Distributed RAM (RAM32M)
www.xilinx.com
Send Feedback
27
Chapter 2: Functional Details
X-Ref Target - Figure 2-8
RAM64X1S
(DI)
D
A[5:0]
6 (D[6:1])
WCLK
WE
6
(CLK)
(WE/CE)
SPRAM64
O
O6
DI1
Output
Registered
Output
D Q
A[6:1]
WA[6:1]
CLK
WE
(Optional)
UG474_c2_07_101210
Figure 2-8:
64 X 1 Single Port Distributed RAM (RAM64X1S)
If four single-port 64 x 1-bit modules are each built as shown in Figure 2-8, the four
RAM64X1S primitives can occupy a SLICEM, as long as they share the same clock, write
enable, and shared read and write port address inputs. This configuration equates to a
64 x 4-bit single-port distributed RAM.
If two dual-port 64 x 1-bit modules are each built as shown in Figure 2-9, the two
RAM64X1D primitives can occupy a SLICEM, as long as they share the same clock, write
enable, and shared read and write port address inputs. This configuration equates to a
64 x 2-bit dual-port distributed RAM.
X-Ref Target - Figure 2-9
RAM64X1D
D
A[5:0]
WCLK
WE
DPRAM64
(DI)
O6
DI1
(D[6:1]) 6
6
(CLK)
(WE/CE)
D Q
A[6:1]
WA[6:1]
CLK
WE
(C[6:1]) 6
6
A[6:1]
WA[6:1]
CLK
WE
Registered
Output
(Optional)
DPRAM64
DI1
DPRA[5:0]
SPO
O6
DPO
D Q
Registered
Output
(Optional)
UG474_c2_08_101210
Figure 2-9:
28
Send Feedback
64 X 1 Dual Port Distributed RAM (RAM64X1D)
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
X-Ref Target - Figure 2-10
RAM64M
DID
ADDRD
WCLK
WE
(DI)
(D[6:1])
(CLK)
(WE)
DPRAM64
O6
DI1
DOD
A[6:1]
WA[6:1]
CLK
WE
D
Registered
Output
Q
(Optional)
DPRAM64
DI1
ADDRC
(C[6:1])
O6
A[6:1]
WA[6:1]
CLK
WE
D
(B[6:1])
O6
D
(A[6:1])
A[6:1]
WA[6:1]
CLK
WE
Registered
Output
Q
(Optional)
DPRAM64
ADDRA
DOB
A[6:1]
WA[6:1]
CLK
WE
DI1
Registered
Output
Q
(Optional)
DPRAM64
DI1
ADDRB
DOC
O6
DOA
D
Registered
Output
Q
(Optional)
UG474_c2_09_070914
Figure 2-10:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
64 X 1 Quad Port Distributed RAM (RAM64M)
www.xilinx.com
Send Feedback
29
Chapter 2: Functional Details
X-Ref Target - Figure 2-11
RAM64M
DPRAM64
unused
unused
WADDR[6:1]
D[6:1] 6
6
(CLK)
(WE)
WCLK
WED
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
DPRAM64
DATA[1]
RADDR[6:1]
C[6:1] 6
6
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[1]
O5
DPRAM64
DATA[2]
B[6:1] 6
6
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[2]
O5
DPRAM64
DATA[3]
A[6:1] 6
6
DI1
DI2
A[6:1]
WA[6:1]
CLK
WE
O6
O[3]
O5
UG474_c2_10_070914
Figure 2-11:
64 X 3 Simple Dual Port Distributed RAM (RAM64M)
Implementation of distributed RAM configurations with depth greater than 64 requires the
usage of wide-function multiplexers (F7AMUX, F7BMUX, and F8MUX), as shown in
Figure 2-12 through Figure 2-14.
If two single-port 128 x 1-bit modules are each built as shown in Figure 2-12, the two
RAM128X1S primitives can occupy a SLICEM, as long as they share the same clock, write
enable, and shared read and write port address inputs. This configuration equates to 128 x
2-bit single-port distributed RAM.
30
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
X-Ref Target - Figure 2-12
RAM128X1S
A6 (CX)
D
SPRAM64
(DI)
O6
DI1
[5:0]
A[6:0]
WCLK
WE
7
(CLK)
(WE/CE)
A[6:1]
WA[7:1]
CLK
WE
0
SPRAM64
DI1
D Q
Registered
Output
(Optional)
[5:0]
7
O6
F7BMUX
Output
A[6:1]
WA[7:1]
CLK
WE
UG474_c2_11_101210
Figure 2-12:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
128 X 1 Single Port Distributed RAM (RAM128X1S)
www.xilinx.com
Send Feedback
31
Chapter 2: Functional Details
X-Ref Target - Figure 2-13
RAM128X1D
A6 (CX)
D
DPRAM64
DI
O6
DI1
6
A[6:0]
7
(CLK)
(WE)
WCLK
WE
A[6:1]
WA[7:1]
CLK
WE
SPO
DPRAM64
Registered
Output
D Q
F7BMUX
O6
DI1
(Optional)
6
7
A[6:1]
WA[7:1]
CLK
WE
DPRAM64
O6
DI1
6
DPRA[6:0]
7
A[6:1]
WA[7:1]
CLK
WE
DPO
DPRAM64
DI1
Registered
Output
D Q
O6
(Optional)
6
7
F7AMUX
A[6:1]
WA[7:1]
CLK
WE
AX
UG474_c2_12_101210
Figure 2-13:
32
Send Feedback
128 X 1 Dual Port Distributed RAM (RAM128X1D)
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Distributed RAM (Available in SLICEM Only)
X-Ref Target - Figure 2-14
RAM256X1S
SPRAM64
D
6
A[7:0]
WCLK
WE
O6
DI1
8
(CLK)
(WE/CE)
A[6:1]
WA[8:1]
CLK
WE
A6 (CX)
SPRAM64
6
8
F7BMUX
O6
DI1
A[6:1]
WA[8:1]
CLK
WE
A7 (BX)
O
F8MUX
SPRAM64
O6
DI1
6
8
A[6:1]
WA[8:1]
CLK
WE
6
8
D Q
Registered
Output
(Optional)
A6 (AX)
SPRAM64
DI1
Output
F7AMUX
O6
A[6:1]
WA[8:1]
CLK
WE
UG474_c2_13_101210
Figure 2-14:
256 X 1 Single Port Distributed RAM (RAM256X1S)
Distributed RAM configurations greater than the provided examples require more than
one SLICEM. There are no direct connections between slices to form larger distributed
RAM configurations within a CLB or between slices.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
33
Chapter 2: Functional Details
Distributed RAM Data Flow
Synchronous Write Operation
The synchronous write operation is a single clock-edge operation with an active-High
write-enable (WE) feature. When WE is High, the input (D) is loaded into the memory
location at address A.
Asynchronous Read Operation
The output is determined by the address A for the single-port mode output SPO of
dual-port mode, or address DPRA for the DPO output of dual-port mode. Each time a new
address is applied to the address pins, the data value in the memory location of that
address is available on the output after the time delay to access the LUT. This operation is
asynchronous and independent of the clock signal.
Distributed RAM Summary
Here is a summary of distributed RAM features:
•
Single-port and dual-port modes are available in SLICEMs
•
A write operation requires one clock edge
•
Read operations are asynchronous (Q output)
•
The data input has a setup-to-clock timing specification
Read Only Memory (ROM)
Each function generator in both SLICEMs and SLICELs can implement a 64 x 1-bit ROM.
Three configurations are available: ROM64X1, ROM128X1, and ROM256X1. ROM contents
are loaded at each device configuration. Table 2-4 shows the number of LUTs occupied by
each ROM configuration size.
Table 2-4:
ROM Configuration
ROM
Number of LUTs
64 x 1
1
128 x 1
2
256 x 1
4
Shift Registers (Available in SLICEM Only)
A SLICEM function generator can also be configured as a 32-bit shift register without using
the flip-flops available in a slice. Used in this way, each LUT can delay serial data from 1 to
32 clock cycles. The shiftin D (DI1 LUT pin) and shiftout Q31 (MC31 LUT pin) lines cascade
LUTs to form larger shift registers. The four LUTs in a SLICEM are thus cascaded to
produce delays up to 128 clock cycles. It is also possible to combine shift registers across
more than one SLICEM. There are no direct connections between slices to form longer shift
registers, nor is the MC31 output at LUT B/C/D available. The resulting programmable
delays can be used to balance the timing of data pipelines.
34
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Shift Registers (Available in SLICEM Only)
Applications for shift registers include:
•
Delay or latency compensation
•
Synchronous FIFO and content addressable memory (CAM)
Shift register functions include:
•
Write operation
•
Synchronous with a clock input (CLK) and an optional clock enable (CE)
•
Fixed read access to Q31
•
Dynamic read access
•
Performed through the 5-bit address bus, A[4:0]
-
•
Any of the 32 bits can be read out asynchronously (at the O6 LUT outputs,
referred to as Q on the primitive) by varying the address
•
This capability is useful in creating smaller shift registers (less than 32 bits).
-
•
•
The LSB of the LUT address is unused and the software automatically ties it
to a logic High.
For example, when building a 13-bit shift register, set the address to the 13th
bit.
A storage element or flip-flop is available to implement a synchronous read.
-
The clock-to-out of the flip-flop determines the overall delay and improves
performance.
-
However, one additional cycle of clock latency is added.
Set or reset of the shift register is not supported.
Figure 2-15 is a logic block diagram of a 32-bit shift register.
X-Ref Target - Figure 2-15
SRLC32E
SHIFTIN (MC31 of Previous LUT)
SRL32
SHIFTIN (D)
(AI)
DI1
MC31
A[4:0]
CLK
CE
5 (A[6:2])
(CLK)
(WE/CE)
SHIFTOUT (Q31)
A[6:2]
CLK
CE
Output (Q)
O6
(AQ) Registered
D Q
Output
(Optional)
UG474_c2_14_110510
Figure 2-15:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
32-Bit Shift Register Configuration
www.xilinx.com
Send Feedback
35
Chapter 2: Functional Details
Figure 2-16 illustrates an example shift register configuration occupying one function
generator.
X-Ref Target - Figure 2-16
32-bit Shift Register
SHIFTIN (D)
WE
SHIFTOUT(Q31)
CLK
5
Address (A[4:0])
MUX
Q
Figure 2-16:
UG474_c2_15_101210
Representation of a Shift Register
Figure 2-17 shows two 16-bit shift registers. The example shown can be implemented in a
single LUT. For more information on the SRLC32E and SRL16E primitives, see UG953,
Vivado Design Suite 7 Series FPGA and Zynq-7000 All Programmable SoC Libraries Guide.
X-Ref Target - Figure 2-17
SRL16
SHIFTIN1 (AI)
A[3:0]
O5
DI1
4
CLK
A[5:2]
CLK
WE
CE
SRL16
DI2
SHIFTIN2 (AX)
4
A[5:2]
CLK
WE
O6
MC31
UG474_c2_16_101210
Figure 2-17:
Dual 16-Bit Shift Register Configuration
As mentioned earlier, the MC31 output and a dedicated connection between shift registers
allows connecting the last bit of one shift register to the first bit of the next, without using
the LUT O6 output. Longer shift registers can be built with dynamic access to any bit in the
chain. The shift register chaining and the F7AMUX, F7BMUX, and F8MUX multiplexers
allow up to a 128-bit shift register with addressable access to be implemented in one
SLICEM. Figure 2-18 through Figure 2-20 illustrate various example shift register
configurations that can occupy one SLICEM.
36
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Shift Registers (Available in SLICEM Only)
X-Ref Target - Figure 2-18
SRL32
SHIFTIN (D)
O6
DI1
5
A[5:0]
MC31
(CLK)
(WE/CE)
CLK
WE
A[6:2]
A5 (AX)
CLK
WE
Output (Q)
SRL32
DI1
F7AMUX
(AQ)
D Q
Registered
Output
O6
(Optional)
5
A[6:2]
CLK
WE
(MC31)
MC31
SHIFTOUT (Q63)
UG474_c2_17_101210
Figure 2-18:
64-Bit Shift Register Configuration
X-Ref Target - Figure 2-19
CX (A5)
SRL32
SHIFTIN (D)
5
A[6:0]
CLK
WE
O6
DI1
A[6:2]
F7BMUX
MC31
(CLK)
(WE/CE)
CLK
WE
BX (A6)
(BMUX)
F8MUX
Registered
Output
(Optional)
SRL32
O6
DI1
5
D Q
Output (Q)
(BQ)
A[6:2]
MC31
CLK
WE
AX (A5)
SRL32
DI1
Not Used
O6
F7AMUX
5
A[6:2]
CLK
WE
UG474_c2_18_101210
Figure 2-19:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
96-Bit Shift Register Configuration
www.xilinx.com
Send Feedback
37
Chapter 2: Functional Details
X-Ref Target - Figure 2-20
SRL32
SHIFTIN (D)
5
A[6:0]
CLK
WE
O6
DI1
(CLK)
(WE/CE)
A[6:2]
MC31
CX (A5)
CLK
WE
SRL32
F7BMUX
O6
DI1
A[6:2]
BX (A6)
MC31
CLK
WE
(BMUX)
F8MUX
D Q
SRL32
(BQ)
Output (Q)
Registered
Output
(Optional)
O6
DI1
A[6:2]
MC31
AX (A5)
CLK
WE
SRL32
O6
DI1
A[6:2]
MC31
CLK
WE
F7AMUX
(MC31)
SHIFTOUT (Q127)
UG474_c2_19_101210
Figure 2-20:
128-Bit Shift Register Configuration
It is possible to create shift registers longer than 128 bits across more than one SLICEM.
However, there are no direct connections between slices to form these shift registers.
Shift Register Data Flow
Shift Operation
Figure 5-7 shows a timing diagram of the shift operation. The shift operation has these
characteristics:
38
•
Operates on a single clock edge
•
Enabled by an active-High clock enable
•
The input (D) is loaded into the first bit of the shift register
•
Each bit is also shifted to the next highest bit position
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Multiplexers
•
In a cascadable shift register configuration, the last bit is shifted out on the MC31
output
•
The bit selected by the 5-bit address port (A[4:0]) appears on the Q output
Dynamic Read Operation
In a dynamic read operation:
•
The Q output is determined by the 5-bit address
•
Each time a new address is applied to the 5-input address pins, the new bit position
value is available on the Q output after the time delay to access the LUT
•
This operation is asynchronous and independent of the clock and clock-enable signals
Static Read Operation
In a static read operation:
•
If the 5-bit address is fixed, the Q output always uses the same bit position
•
This mode implements any shift-register length from 1 to 32 bits in one LUT
•
The shift register length is (N + 1), where N is the input address (0–31)
•
The Q output changes synchronously with each shift operation
•
The previous bit is shifted to the next position and appears on the Q output
Shift Register Summary
Here is a summary of the shift registers:
•
A shift operation requires one clock edge
•
Dynamic-length read operations to the Q output of the LUT are asynchronous
•
Static-length read operations to the Q output of the LUT are synchronous
•
The data input has a setup-to-clock timing specification
•
In a cascadable configuration, the Q31 output always contains the last bit value
•
The Q31 output changes synchronously after each shift operation
Multiplexers
Function generators and associated multiplexers in 7 series FPGAs can implement these
functions:
•
4:1 multiplexers using one LUT
•
•
8:1 multiplexers using two LUTs
•
•
Four 4:1 MUXes per slice
Two 8:1 MUXes per slice
16:1 multiplexers using four LUTs
•
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
One 16:1 MUX per slice
www.xilinx.com
Send Feedback
39
Chapter 2: Functional Details
These wide input multiplexers are implemented in one level of logic (or LUT) using the
dedicated F7AMUX, F7BMUX, and F8MUX multiplexers. These multiplexers allow LUT
combinations of up to four LUTs in a slice. The wide multiplexers can also be used to create
general-purpose functions of up to 13 inputs in two LUTs or 27 inputs in four LUTs (one
slice). Although the multiplexer outputs are combinatorial, they can be registered in the
CLB storage elements.
Designing Large Multiplexers
4:1 Multiplexer
Each LUT can be configured into a 4:1 MUX. The 4:1 MUX can be implemented with a
flip-flop in the same slice. Up to four 4:1 MUXes can be implemented in a slice, as shown in
Figure 2-21.
X-Ref Target - Figure 2-21
SLICE
LUT
(D)
O6
SEL D [1:0], DATA D [3:0]
Input
(D[6:1]) 6
D Q
A[6:1]
(DQ)
(C)
O6
(C[6:1]) 6
D Q
A[6:1]
(CQ)
(B)
O6
(B[6:1]) 6
D Q
A[6:1]
6
A[6:1]
(CLK)
(BQ)
(A)
O6
SEL A [1:0], DATA A [3:0]
Input
Registered
Output
4:1 MUX Output
Registered
Output
(Optional)
LUT
(A[6:1])
4:1 MUX Output
(Optional)
LUT
SEL B [1:0], DATA B [3:0]
Input
Registered
Output
(Optional)
LUT
SEL C [1:0], DATA C [3:0]
Input
4:1 MUX Output
D Q
(AQ)
4:1 MUX Output
Registered
Output
(Optional)
CLK
UG474_c2_20_101210
Figure 2-21:
40
Send Feedback
Four 4:1 Multiplexers in a Slice
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Multiplexers
8:1 Multiplexer
Each slice has an F7AMUX and an F7BMUX. These two MUXes combine the output of two
LUTs to form a combinatorial function up to 13 inputs (or an 8:1 MUX). Up to two 8:1
MUXes can be implemented in a slice, as shown in Figure 2-22.
X-Ref Target - Figure 2-22
SLICE
LUT
O6
SEL D [1:0], DATA D [3:0]
Input (1)
(D[6:1]) 6
A[6:1]
F7BMUX
(CMUX)
LUT
O6
SEL C [1:0], DATA C [3:0]
Input (1)
SELF7(1)
CLK
(C[6:1]) 6
D Q
(CQ)
8:1 MUX
Output (1)
Registered
Output
A[6:1]
(Optional)
(CX)
(CLK)
LUT
O6
SEL B [1:0], DATA B [3:0]
Input (2)
(B[6:1]) 6
A[6:1]
F7AMUX
(AMUX)
LUT
O6
SEL A [1:0], DATA A [3:0]
Input (2)
SELF7(2)
(A[6:1])
6
D Q
(AQ)
8:1 MUX
Output (2)
Registered
Output
A[6:1]
(Optional)
(AX)
UG474_c2_21_101210
Figure 2-22:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Two 8:1 Multiplexers in a Slice
www.xilinx.com
Send Feedback
41
Chapter 2: Functional Details
16:1 Multiplexer
Each slice has an F8MUX. F8MUX combines the outputs of F7AMUX and F7BMUX to form
a combinatorial function up to 27 inputs (or a 16:1 MUX). Only one 16:1 MUX can be
implemented in a slice, as shown in Figure 2-23.
X-Ref Target - Figure 2-23
SLICE
LUT
O6
SEL D [1:0], DATA D [3:0]
Input
(D[6:1]) 6
F7BMUX
A[6:1]
LUT
O6
SEL C [1:0], DATA C [3:0]
Input
SELF7
(C[6:1]) 6
A[6:1]
F8MUX
(CX)
(BMUX)
LUT
D Q
O6
SEL B [1:0], DATA B [3:0]
Input
(B[6:1]) 6
F7AMUX
A[6:1]
(B)
16:1 MUX
Output
Registered
Output
(Optional)
LUT
O6
SEL A [1:0], DATA A [3:0]
Input
SELF7
SELF8
CLK
(A[6:1])
6
A[6:1]
(AX)
(BX)
(CLK)
UG474_c2_22_101210
Figure 2-23:
16:1 Multiplexer in a Slice
It is possible to create multiplexers wider than 16:1 across more than one SLICEM.
However, there are no direct connections between slices to form these wide multiplexers.
42
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Carry Logic
Carry Logic
In addition to function generators, dedicated fast lookahead carry logic is provided to
perform fast arithmetic addition and subtraction in a slice. A 7 series FPGA CLB has two
separate carry chains, as shown in Figure 1-1. The carry chains are cascadable to form
wider add/subtract logic, as shown in Figure 2-2.
The carry chain runs upward and has a height of four bits per slice. For each bit, there is a
carry multiplexer (MUXCY) and a dedicated XOR gate for adding/subtracting the
operands with a selected carry bits. The dedicated carry path and carry multiplexer
(MUXCY) can also be used to cascade function generators for implementing wide logic
functions.
Figure 2-24 illustrates the carry chain with associated logic in a slice.
X-Ref Target - Figure 2-24
COUT (To Next Slice)
Carry Chain Block
(CARRY4)
CO3
O6 From LUTD
S3
DMUX/DQ*
MUXCY
O3
O5 From LUTD
DMUX
DI3
D Q
DX
DQ
(Optional)
CO2
O6 From LUTC
S2
CMUX/CQ*
MUXCY
O2
O5 From LUTC
CMUX
DI2
D Q
CX
CQ
(Optional)
CO1
O6 From LUTB
S1
BMUX/BQ*
MUXCY
O1
O5 From LUTB
BMUX
DI1
D Q
BX
BQ
(Optional)
CO0
O6 From LUTA
S0
AMUX/AQ*
MUXCY
O0
O5 From LUTA
AMUX
DI0
D Q
AX
CYINIT
CIN
01
AQ
(Optional)
* Can be used if
unregistered/registered
outputs are free.
CIN (From Previous Slice)
UG474_c2_23_071813
Figure 2-24:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Fast Carry Logic Path and Associated Elements
www.xilinx.com
Send Feedback
43
Chapter 2: Functional Details
The carry chains use carry lookahead logic along with the function generators. There are
10 independent inputs and 8 independent outputs:
•
Inputs
•
•
S inputs - S0 to S3
-
The “propagate” signals of the carry lookahead logic
-
Sourced from the O6 output of a function generator
DI inputs - DI1 to DI4
-
The “generate” signals of the carry lookahead logic
-
Sourced from either the O5 output of a function generator
- To create a multiplier
-
Or the BYPASS input (AX, BX, CX, or DX) of a slice
- To create an adder/accumulator
•
CYINIT
-
CIN of the first bit in a carry chain
- 0 for add
- 1 for subtract
- AX input for the dynamic first carry bit
•
CIN
-
•
Cascade slices to form a longer carry chain
Outputs
•
O outputs - O0 to O3
-
•
Contain the sum of the addition/subtraction
CO outputs - CO0 to CO3
-
Compute the carry out for each bit
-
CO3 is connected to COUT output of a slice to form a longer carry chain by
cascading multiple slices
The propagation delay for an adder increases linearly with the number of bits in the
operand, as more carry chains are cascaded. The carry chain can be implemented with a
storage element or a flip-flop in the same slice.
Carry logic cascading is limited only by the height of the column of slices. Carry logic
cannot be cascaded across super logic regions (SLRs) in devices using stacked silicon
interconnect (SSI) technology. See Devices Using Stacked Silicon Interconnect (SSI)
Technology in Chapter 6.
44
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 3
Design Entry
CLB resources are used automatically and efficiently by FPGA synthesis tools without any
special FPGA-specific coding required. Some HDL coding suggestions and techniques can
help optimize designs for maximum efficiency.
Design Checklist
These guidelines are provided as a quick checklist of design suggestions for effective use of
the 7 series CLBs:
•
Resource Utilization
•
Use generic HDL code and allow the synthesis and mapping tools to choose the
specific FPGA CLB resources.
•
Consider instantiation of specific resources only when necessary to meet density
or performance requirements.
•
Compare the results to an estimated slice count to verify design efficiency.
•
If a design is running out of resources in the target device, examine which
resource is the limiting factor and consider using alternative resources, such as
moving registers to SRLs or distributed RAM, or distributed RAM to block RAM,
or carry logic to DSP slices.
•
When migrating a design from another architecture, remove resource
instantiation and any mapping and floorplanning constraints (refer to UG429,
7 Series FPGAs Migration Methodology Guide).
Refer to www.xilinx.com for good HDL coding techniques for FPGAs, such as in UG687,
XST User Guide for Virtex-6, Spartan-6, and 7 Series Devices.
•
Pipelining
•
•
The designer should use sequential design techniques and pipelining to take
advantage of abundant flip-flops for performance.
Control Signals
•
Use control signals only as necessary.
•
Avoid using a routed global reset signal and minimize use of local resets to
maximize opportunity to use FPGA resources.
•
Use active-High control signals.
•
Avoid having both set and reset on the same flip-flop.
•
Avoid control signals on small shift registers and storage arrays in order to use
LUTs instead of flip-flops, to maximize utilization and minimize power.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
45
Chapter 3: Design Entry
•
Software Options
•
To improve performance automatically, use timing constraints and trade off
longer implementation run times through software options.
Using the CLB Resources
Xilinx recommends using generic HDL code and allowing the tools to infer the usage of
CLB resources. Using IP solutions designed for 7 series FPGAs can help make full use of
the CLB resources. Although any feature in the CLB can be instantiated directly, including
the LUTs, carry logic, and sequential elements, instantiation should be reserved primarily
for specifying when resources outside the CLB should be used, such as the DSP slice.
Instantiation might also be needed if the synthesis tool does not infer a desired special CLB
resource, such as the wide multiplexers, distributed RAM, or SRL feature.
Primitives
This section provides an overview of the most commonly used CLB primitives. For
instantiation examples and for information on other CLB primitives, see UG953, Vivado
Design Suite 7 Series FPGA and Zynq-7000 All Programmable SoC Libraries Guide.
Multiplexer Primitives
The multiplexer primitives provide direct instantiation of the dedicated multiplexers in
each slice, allowing wider multiplexers to be built. Table 3-1 describes the two primitives.
Table 3-1:
Multiplexer Primitives
Primitive
Inputs
Resource
Output Function
MUXF7
LUT outputs
(4:1 multiplexer)
F7AMUX or F7BMUX
8:1 multiplexer
MUXF8
F7AMUX and F7BMUX outputs
(8:1 multiplexer)
F8MUX
16:1 multiplexer
The port signals are the same for both multiplexer primitives. Figure 3-1 shows MUXF7.
X-Ref Target - Figure 3-1
MUXF7
I0
O
I1
S
UG474_c3_01_101210
Figure 3-1:
46
Send Feedback
www.xilinx.com
MUXF7 Primitive
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Primitives
Port Signals
Data In - I0, I1
The data input provides the data to be selected by the select signal (S).
Control In - S
The select input signal determines the data input signal to be connected to the output O.
Logic 0 selects the I0 input, while logic 1 selects the I1 input.
Data Out - O
The data output O provides the data value (one bit) selected by the control inputs.
Carry Chain Primitive
The CARRY4 primitive instantiates the fast carry logic available in each slice. This
primitive works in conjunction with LUTs to build adders and multipliers. Figure 3-2
shows the CARRY4 primitive. Synthesis tools generally infer this logic from arithmetic
HDL code, automatically connecting this function properly.
X-Ref Target - Figure 3-2
CARRY4
DI[3:0]
O[3:0]
CO[3:0]
S[3:0]
CYINIT
CI
UG474_c3_02_101210
Figure 3-2:
CARRY4 Primitive
Port Signals
Sum Outputs - O[3:0]
The sum outputs provide the final result of the addition/subtraction. They connect to the
slice AMUX/BMUX/CMUX/DMUX outputs.
Carry Outputs - CO[3:0]
The carry outputs provide the carry out for each bit. CO[3] is equivalent to COUT. A longer
carry chain can be created if CO[3] is connected through COUT to the CI input of another
CARRY4 primitive, and dedicated routing connects the carry chain up a column of slices.
The carry outputs also optionally connect to the slice AMUX/BMUX/CMUX/DMUX
outputs.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
47
Chapter 3: Design Entry
Carry In - CI
The carry in input, also called CIN, is used to cascade slices to form longer carry chains.
Data Inputs - DI[3:0]
The data inputs are used as “generate” signals to the carry lookahead logic. The “generate”
signals are sourced from LUT outputs.
Select Inputs - S[3:0]
The select inputs are used as “propagate” signals to the carry lookahead logic. The
“propagate” signals are sourced from LUT outputs.
Carry Initialize - CYINIT
The carry initialize input is used to select the first bit in a carry chain. The value for this pin
is either 0 (for add), 1 (for subtract), or AX input (for the dynamic first carry bit).
SLICEM Distributed RAM Primitives
Eight primitives are shown in Table 3-2. Three primitives are single-port RAM, three
primitives are dual-port RAM, and two primitives are quad-port RAM.
Table 3-2:
Single-Port, Dual-Port, and Quad-Port Distributed RAM
Primitive
RAM Size
Type
RAM32X1S
32-bit
Single-port
RAM32M
32-bit
Quad-port
Address Inputs
A[4:0] (read/write)
ADDRA[4:0] (read)
ADDRB[4:0] (read)
ADDRC[4:0] (read)
ADDRD[4:0] (read/write)
RAM64X1S
64-bit
Single-port
A[5:0] (read/write)
RAM64X1D
64-bit
Dual-port
A[5:0] (read/write)
DPRA[5:0] (read)
RAM64M
64-bit
Quad-port
ADDRA[5:0] (read)
ADDRB[5:0] (read)
ADDRC[5:0] (read)
ADDRD[5:0] (read/write)
RAM128X1S
128-bit
Single-port
A[6:0] (read/write)
RAM128X1D
128-bit
Dual-port
A[6:0], (read/write)
DPRA[6:0] (read)
RAM256X1S
256-bit
Single-port
A[7:0] (read/write)
The input and output data are one bit wide (with the exception of the quad-port RAM).
Figure 3-3 shows generic single-port, dual-port, and quad-port distributed RAM
primitives. The A, ADDR, and DPRA signals are address buses.
48
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Primitives
X-Ref Target - Figure 3-3
RAM#X1S
D
WE
RAM#X1D
O
D
WE
RAM#M
SPO
WCLK
WCLK
A[#:0]
A[#:0]
R/W Port
DPRA[#:0]
Read Port
DI[A:D][#:0]
WE
DOD[#:0]
WCLK
DPO
ADDRD[#:0]
R/W Port
ADDRC[#:0]
Read Port
DOC[#:0]
ADDRB[#:0]
Read Port
DOB[#:0]
ADDRA[#:0]
Read Port
DOA[#:0]
UG474_c3_03_101210
Figure 3-3:
Single-Port, Dual-Port, and Quad-Port Distributed RAM Primitives
Instantiating several distributed RAM primitives can be used to implement wide memory
blocks.
Port Signals
Each distributed RAM port operates independently of the other while reading the same set
of memory cells.
Clock – WCLK
The clock is used for the synchronous write. The data and the address input pins have
setup times referenced to the WCLK pin.
The clock pin (WCLK) has an inversion option at the slice level. The clock signal can be
active at the negative edge of the clock or the positive edge of the clock without requiring
other logic resources. The default is at the positive clock edge.
Enable – WE/WED
The enable pin affects the write functionality of the port. An inactive write enable prevents
any writing to memory cells. An active write enable causes the clock edge to write the data
input signal to the memory location pointed to by the address inputs.
Address – A[#:0], DPRA[#:0], and ADDRA[#:0] – ADDRD[#:0]
The address inputs A[#:0] (for single-port and dual-port), DPRA[#:0] (for dual-port), and
ADDRA[#:0] – ADDRD[#:0] (for quad-port) select the memory cells for read or write. The
width of the port determines the required address inputs. Some of the address inputs are
not buses in VHDL or Verilog instantiations. Table 3-2 summarizes the function of each
address pin.
Data In – D, DID[#:0]
The data input D (for single-port and dual-port) and DID[#:0] (for quad-port) provide the
new data value to be written into the RAM.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
49
Chapter 3: Design Entry
Data Out – O, SPO, DPO and DOA[#:0] – DOD[#:0]
The data out O (single-port or SPO), DPO (dual-port), and DOA[#:0] – DOD[#:0]
(quad-port) reflects the contents of the memory cells referenced by the address inputs.
Following an active write clock edge, the data out (O, SPO, or DOD[#:0]) reflects the newly
written data.
SLICEM SRL Shift Register Primitive
One primitive is available for the 32-bit shift register (SRLC32E), which uses one LUT in a
SLICEM. Figure 3-4 shows the 32-bit shift register primitive.
X-Ref Target - Figure 3-4
SRLC32E
5
D
A[4:0]
Q
CE
Q31
CLK
UG474_c3_04_102910
Figure 3-4:
32-Bit Shift Register
Port Signals
Clock – CLK
Either the rising edge or the falling edge of the clock is used for the synchronous shift
operation. The data and clock enable input pins have setup times referenced to the chosen
edge of CLK.
The clock pin (CLK) has an inversion option at the slice level. The clock signal can be active
at the negative or positive edge without requiring other logic resources. The default is
positive clock edge.
Data In – D
The data input provides new data (one bit) to be shifted into the shift register.
Clock Enable – CE
The clock enable pin affects shift functionality. An inactive clock enable pin does not shift
data into the shift register and does not write new data. Activating the clock enable allows
the data in (D) to be written to the first location and all data to be shifted by one location.
When available, new data appears on output pins (Q) and the cascadable output pin (Q31).
Address – A[4:0]
The address input selects the bit (range 0 to 31) to be read. The nth bit is available on the
output pin (Q). Address inputs have no effect on the cascadable output pin (Q31). Q31 is
always the last bit of the shift register (bit 31).
Data Out – Q
The data output Q provides the data value (1 bit) selected by the address inputs.
50
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Primitives
Data Out – Q31
The data output Q31 provides the last bit value of the 32-bit shift register. New data
becomes available after each shift-in operation.
Flip-Flop Primitives
There are several primitives for the CLB storage elements, including both flip-flops and
latches, with different combinations of control signals available. The FDRE primitive is
shown in Figure 3-5 for an example. For more information on the flip-flop and latch
primitives, see UG953, Vivado Design Suite 7 Series FPGA and Zynq-7000 All Programmable
SoC Libraries Guide.
X-Ref Target - Figure 3-5
FDRE
D
Q
CE
C
R
UG474_c3_05_102910
Figure 3-5:
FDRE Primitive
Port Signals
Data In - D
The data input provides new data (one bit) to be clocked into the flip-flop.
Data Out - Q
Q is the one-bit registered data output from the flip-flop.
Clock - C
Either the rising edge or the falling edge of the clock is used to capture data and toggle the
output. The data and clock enable input pins have setup times referenced to the chosen
edge of the clock. The clock pin (C) has an inversion option at the slice level. The clock
signal can be active at the negative or positive edge of the clock without requiring other
logic resources. The default is the positive clock edge. All flip-flops in a slice must use the
same clock and same clock polarity.
Clock Enable - CE
When CE is Low, clock transitions are ignored with respect to the D input. The set and reset
inputs have priority over the clock enable.
Synchronous Reset - R
When R is High, all other inputs are overridden and the data output (Q) is driven Low on
the active clock transition. This signal is available in the FDRE component. The FDRE
flip-flop is also cleared by default on power-up.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
51
Chapter 3: Design Entry
Synchronous Set - S
When S is High, all other inputs are overridden and the data output (Q) is driven High on
the active clock transition. This signal is available in the FDSE component. The FDSE
flip-flop is also preset by default on power-up.
Asynchronous Clear - CLR
When CLR is High, all other inputs are overridden and the data output (Q) is driven Low.
This signal is available in the FDCE component. The FDCE flip-flop is also cleared by
default on power-up.
Asynchronous Preset - PRE
When PRE is High, all other inputs are overridden and the data output (Q) is driven High.
This signal is available in the FDPE component. The FDPE flip-flop is also preset by default
on power-up.
Note: Using both asynchronous clear and preset on the same flip-flop requires additional resources
and timing paths.
52
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 4
Applications
This chapter provides guidance regarding using the CLB resources as parts of larger
functions. Trade-offs are described between alternative methods of implementing these
functions.
Distributed RAM Applications
Distributed RAM provides a trade-off between using storage elements for very small
arrays and block RAM for larger arrays. It is recommended to infer memory where
possible to provide the greatest flexibility. Distributed RAM can also be targeted by
instantiation or through the use of Xilinx LogiCORE™ IP.
In general, distributed RAM should be used for all memories that consist of 64 bits or less,
unless there is a shortage of SLICEM or logic resources for the target device. Distributed
RAM is more efficient in terms of resources, performance, and power.
For depths greater than 64 bits but less than or equal to 128 bits, the decision on the best
resource to use depends on these factors:
1.
The availability of extra block RAMs. If not available, distributed RAM should be
used.
2.
The latency requirements. If asynchronous read capability is needed, distributed
RAMs must be used.
3.
The data width. Widths greater than 16 bits should use block RAM, if available.
4.
The necessary performance requirements. Registered distributed RAMs generally
have shorter clock-to-out timing and fewer placement restrictions than block RAMs.
Shift Register Applications
Synchronous Shift Registers
The shift-register primitive does not use the register available in the same slice. To
implement a fully synchronous read and write shift register, output pin Q must be
connected to a flip-flop. Both the shift register and the flip-flop share the same clock, as
shown in Figure 4-1.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
53
Chapter 4: Applications
X-Ref Target - Figure 4-1
SRLC32E
D
FF
D
Q
Synchronous
Output
Q
Address
(Write Enable)
CE
Q31
CLK
UG474_c4_01_101210
Figure 4-1:
Synchronous Shift Register
This configuration provides a better timing solution and simplifies the design. Because the
flip-flop must be considered to be the last register in the shift-register chain, the static or
dynamic address should point to the desired length minus one. If needed, the cascadable
output can also be registered in a flip-flop.
Static-Length Shift Registers
The cascadable 32-bit shift register implements any static length mode shift register
without the dedicated multiplexers (F7AMUX, F7BMUX, and F8MUX). Figure 4-2
illustrates a 72-bit shift register. Only the last SRLC32E primitive needs to have its address
inputs tied to 0b00111. Alternatively, shift register length can be limited to 71 bits
(address tied to 0b00110) and a flip-flop can be used as the last register. (In an SRLC32E
primitive, the shift register length is the address input + 1).
X-Ref Target - Figure 4-2
LUT
D
LUT
D
D
D
Q31
Q31
SRLC32E
SRLC32E
LUT
LUT
D
D
Q31
Q31
SRLC32E
SRLC32E
FF
LUT
D
00111
5
LUT
Q
A[4:0]
OUT
(72-bit SRL)
D
00110
5
Q
D
A[4:0]
Q31
Q31
SRLC32E
SRLC32E
Q
OUT
(72-bit SRL)
UG474_c4_02_101210
Figure 4-2: Example Static-Length Shift Register
54
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Carry Logic Applications
Carry Logic Applications
Dedicated carry logic in the CLB improves the performance of arithmetic functions such as
adders, counters, and comparators. Designs that include simple counters or adder/
subtractors automatically infer the carry logic. More complex functions including
multipliers can be implemented using the separate DSP48E1 slice.
The DSP48E1 slices in every device support many independent functions including
multiply, multiply accumulate, multiply add, three-input add, barrel shift, wide-bus
multiplexing, magnitude comparator, bitwise logic functions, pattern detection, and wide
counter. The architecture also supports cascading multiple DSP48E1 slices to form wide
math functions, DSP filters, and complex arithmetic. For more details on the DSP48E1 slice,
see UG479, 7 Series DSP48E1 Slice User Guide.
The choice between using CLB carry logic and the DSP48E1 slice depends on the
application. A small arithmetic function can be faster and lower power using the CLB carry
logic rather than using an entire DSP48E1 slice. When the DSP48E1 slices are fully utilized,
the CLB slices and carry logic provide an efficient alternative.
Using Carry Logic
Carry logic can be inferred or instantiated. Using macros designed for the 7 series FPGA
can provide the most flexibility and efficiency, especially for more complex functions.
Unimacros include adders, counters, comparators, multipliers, and multiplier/
accumulators. The unimacros use the DSP48E1 slice.
The Xilinx IP includes a similar set of functions. When defining these functions, the user
can specify whether the implementation should be in the CLB carry logic or in the
DSP48E1 slice.
The carry logic runs vertically up every other column of slices (SLICEL and SLICEM). The
Xilinx tools automatically place logic in a column when the carry logic is used. When
floorplanning, carry logic based functions should always be specified as a group to avoid
unnecessary breaks in the carry chain. Carry logic cannot be cascaded across super logic
regions (SLRs) in devices using stacked silicon interconnect (SSI) technology.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
55
Chapter 4: Applications
56
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 5
Timing
Although it is not necessary to understand the various timing parameters to implement
most designs using Xilinx software, a thorough timing model can assist advanced users in
analyzing critical paths or planning speed-sensitive designs, especially in the larger and
more complex 7 series FPGAs.
The user is recommended to use the models in this chapter in conjunction with both the
Xilinx® Timing Analyzer and the section on switching characteristics in the respective
7 series FPGAs data sheet. All pin names, parameter names, and paths are consistent with
the Timing Analyzer reports. Most of the CLB timing parameters found in the data sheet
section on switching characteristics are described in this chapter.
Five categories of functions are described in this chapter:
•
CLB General Slice Timing Model and Parameters, page 58
•
CLB Slice Multiplexer Timing Model and Parameters, page 60
•
CLB Slice Carry-Chain Timing Model and Parameters, page 61
•
CLB Slice Distributed RAM Timing Model and Parameters (Available in SLICEM
Only), page 63
•
CLB Slice SRL Shift Register Timing Model and Parameters (Available in SLICEM
Only), page 66
For each category, three timing model sections are described:
•
Functional element diagram - basic architectural schematic illustrating pins and
connections
•
Timing parameters - definitions of the respective 7 series FPGAs data sheet timing
parameters
•
Timing diagram - illustrates functional element timing parameters relative to each
other
The general form of a timing parameter name is T subscripted with a combination of the
starting pin and the ending pin. For example, TCECK is the setup time from CE to the clock,
and TCKCE is the hold time from the clock to CE being removed. In some cases, a single
parameter name can apply to multiple paths through the slice, and can have different
values for each path.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
57
Chapter 5: Timing
CLB General Slice Timing Model and Parameters
A simplified 7 series FPGA slice is shown in Figure 5-1. Only the elements relevant to the
timing paths described in this section are shown.
X-Ref Target - Figure 5-1
LUT
D
Inputs
6
O6
D
O5
DMUX
FF/LAT
D
CE
CLK
DX
Q
D
6
DQ
SR
CE
CK
LUT
C
Inputs
Q
SR
F7BMUX
O6
C
O5
CMUX
FF/LAT
CX
Q
D
F8MUX
6
Q
CQ
SR
CE
CK
SR
LUT
B
Inputs
D
CE
CLK
O6
B
BMUX
O5
FF/LAT
BX
Q
D
LUT
A
Inputs
6
F7AMUX
D
CE
CLK
Q
BQ
SR
CE
CK
SR
O6
A
O5
AMUX
AX
Q
D
FF/LAT
CE
CK
SR
CE
D
CE
CLK
Q
AQ
SR
CLK
SR
UG474_c5_01_101210
Figure 5-1:
58
Send Feedback
Simplified 7 Series FPGA Slice
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB General Slice Timing Model and Parameters
General Timing Parameters
Table 5-1 shows the general slice timing parameters for a majority of the paths in
Figure 5-1.
Table 5-1:
General Slice Timing Parameters
Parameter
Function
Description
TILO(1)
A/B/C/D inputs to A/B/C/D
outputs
Propagation delay from the A/B/C/D LUT inputs of the slice,
through the look-up tables (LUTs), to the A/B/C/D outputs of the
slice.
TITO
A/B/C/D inputs through
transparent latch to AQ/BQ/
CQ/DQ outputs
Propagation delay from the A/B/C/D LUT inputs of the slice,
through the LUTs to the AQ/BQ/CQ/DQ outputs of the slice
sequential elements (configured as a latch).
TCKO
FF Clock (CLK) to AQ/BQ/
CQ/DQ outputs
Time after the clock that data is stable at the AQ/BQ/CQ/DQ
outputs of the slice sequential elements (configured as a flip-flop).
TCKLO
Latch Clock (CLK) to AQ/BQ/
CQ/DQ outputs
Time after the clock that data is stable at the AQ/BQ/CQ/DQ
outputs of the slice sequential elements (configured as a latch).
TSHCKO
FF Clock (CLK) to AMUX/
BMUX/CMUX/DMUX
outputs
Time after the clock that data is stable at the AMUX/BMUX/
CMUX/DMUX outputs (through the slice flip-flops).
Combinatorial Delays
Sequential Delays
Setup and Hold Times for Slice Sequential Elements(2)
TAS/TAH
A/B/C/D inputs
Time before/after the CLK that data from the A/B/C/D LUT
inputs of the slice must be stable.
TDICK/TCKDI
AX/BX/CX/DX inputs
Time before/after the CLK that data from the AX/BX/CX/DX
inputs of the slice must be stable.
TCECK/TCKCE
CE input
Time before/after the CLK that the CE input of the slice must be
stable.
TSRCK/TCKSR
SR input
Time before/after the CLK that the SR (Set/Reset) input of the
slice must be stable.
TRPW (TSRMIN)
SR input
Minimum pulse width for the SR (Set/Reset).
TRQ
SR input
Propagation delay for an asynchronous Set/Reset of the slice
sequential elements.
TCEO
CE input
Propagation delay from CE enable to latch output at AQ/BQ/
CQ/DQ outputs.
FTOG
CLK input
Toggle Frequency – Maximum frequency that a slice flip-flop can
be clocked. For export control.
Set/Reset
Miscellaneous
Notes:
1. This parameter includes a LUT configured as two five-input functions.
2. TXXCK = Setup Time (before clock edge), and TCKXX = Hold Time (after clock edge).
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
59
Chapter 5: Timing
General Timing Characteristics
Figure 5-2 illustrates the general timing characteristics of a 7 series FPGA slice.
X-Ref Target - Figure 5-2
1
2
3
CLK
TCECK
CE
TDICK
AX/BX/CX/DX
(DATA)
TSRCK
SR (RESET)
TCKO
TCKO
AQ/BQ/CQ/DQ
(OUT)
UG474_c5_02_102310
Figure 5-2:
General Slice Timing Characteristics
•
At time TCECK before clock event (1), the clock-enable signal becomes valid-High at
the CE input of the slice register.
•
At time TDICK before clock event (1), data from either AX, BX, CX, or DX inputs
become valid-High at the D input of the slice register and is reflected on either the
AQ, BQ, CQ, or DQ pin at time TCKO after clock event (1).
•
At time TSRCK before clock event (3), the SR signal (configured as synchronous reset)
becomes valid-High, resetting the slice register. This is reflected on the AQ, BQ, CQ,
or DQ pin at time TCKO after clock event (3).
CLB Slice Multiplexer Timing Model and Parameters
Figure 2-22, page 41 and Figure 2-23, page 42 illustrate the wide multiplexers in a 7 series
FPGA slice.
Slice Multiplexer Timing Parameters
Table 5-2 shows the slice multiplexer timing parameters for a majority of the paths in
Figure 2-22 and Figure 2-23. Note that many of these parameter names are also used for
paths that do not include the wide multiplexers.
Table 5-2:
Slice Multiplexer Timing Parameters
Parameter
Function
Description
Propagation Delays for Slice Using the Wide Multiplexers
TAXA/TAXB/
TBXB/
TCXB/TCXC
60
Send Feedback
AX/BX/CX inputs to AMUX/
BMUX/CMUX outputs
www.xilinx.com
Propagation delay from the
AX/BX/CX inputs of the
slice through the select inputs
of the multiplexers to the
AMUX/BMUX/CMUX
outputs of the slice.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Slice Carry-Chain Timing Model and Parameters
Table 5-2:
Slice Multiplexer Timing Parameters (Cont’d)
Parameter
Function
Description
TOPAB/
TOPBA/TOPBB/
TOPCB/
TOPDB/ TOPDC
A/B/C/D inputs to AMUX/
BMUX/CMUX outputs
Propagation delay from the
A/B/C/D LUT inputs of the
slice through the
multiplexers to the AMUX/
BMUX/CMUX outputs of
the slice.
TILO_2
A/C inputs to AMUX/CMUX
outputs
Propagation delay from the A
LUT inputs of the slice
through the F7AMUX
multiplexer to the AMUX
output of the slice, or
propagation delay from the C
LUT inputs of the slice
through the F7BMUX
multiplexer to the CMUX
output of the slice.
CLB Slice Carry-Chain Timing Model and Parameters
Figure 2-24, page 43 illustrates a carry chain in a 7 series FPGA slice.
Slice Carry-Chain Timing Parameters
Table 5-3 shows the slice carry-chain timing parameters for a majority of the paths in
Figure 2-24.
Table 5-3:
Slice Carry-Chain Timing Parameters
Parameter
Function
Description
Propagation Delays for Slice Configured as a Carry Chain
TAXA/TAXB/TAXC/TAXD/
TBXB/TBXC/TBXD/
TCXC/TCXD/
TDXD
AX/BX/CX/DX inputs to
AMUX/BMUX/ CMUX/
DMUX outputs
Propagation delay from the
AX/BX/CX/DX inputs of
the slice through the carry
logic to the AMUX/BMUX/
CMUX/DMUX outputs of
the slice.
TAXCY/TBXCY/TCXCY/TDXCY
AX/BX/CX/DX inputs to
COUT output
Propagation delay from the
AX/BX/CX/DX inputs of
the slice to the COUT output
of the slice.
TBYP
CIN input to COUT output
Propagation delay from the
CIN input of the slice to the
COUT output of the slice.
TOPCYA/TOPCYB/TOPCYC/
TOPCYD
A/B/C/D inputs to COUT
output
Propagation delay from the
A/B/C/D LUT inputs of the
slice to the COUT output of
the slice.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
61
Chapter 5: Timing
Table 5-3:
Slice Carry-Chain Timing Parameters (Cont’d)
Parameter
Function
TCINA/TCINB/TCINC/TCIND
Description
CIN input to AMUX/
BMUX/ CMUX/DMUX
outputs
Propagation delay from the
CIN input of the slice to
AMUX/BMUX/CMUX/
DMUX outputs of the slice
using XOR (sum).
Setup and Hold Times for a Slice Configured as a Carry Chain(1)
TCINCK/TCKCIN
CIN input
Time before/after the CLK
that the CIN input of the slice
must be stable.
Notes:
1. TXXCK = Setup Time (before clock edge), and TCKXX = Hold Time (after clock edge).
Slice Carry-Chain Timing Characteristics
Figure 5-3 illustrates the timing characteristics of a slice carry chain implemented in a
7 series FPGA slice.
X-Ref Target - Figure 5-3
1
CLK
CIN
(DATA)
TCINCK
TCKO
AQ/BQ/CQ/DQ
(OUT)
UG474_c5_03_110110
Figure 5-3:
Slice Carry-Chain Timing Characteristics
Note relevant to Figure 5-3:
•
62
At time TCINCK before clock event 1, data from the CIN input becomes valid-High at
the Data input of the slice register. This is reflected on any of the AQ/BQ/CQ/DQ
pins at time TCKO after clock event 1.
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Slice Distributed RAM Timing Model and Parameters (Available in SLICEM Only)
CLB Slice Distributed RAM Timing Model and Parameters
(Available in SLICEM Only)
Figure 5-4 illustrates the details of distributed RAM implemented in a 7 series FPGA slice.
Some elements of the slice are omitted for clarity. Only the elements relevant to the timing
paths described in this section are shown.
X-Ref Target - Figure 5-4
RAM
DI
DX
D input
6
CLK
WE
DI1
DI2
A[6:0]
WA[6:0]
CLK
WE
O6
D
O5
DMUX
O6
C
O5
CMUX
O6
B
O5
BMUX
O6
A
O5
AMUX
RAM
CI
CX
C input
6
DI1
DI2
A[6:0]
WA[6:0]
CLK
WE
RAM
BI
BX
B input
6
DI1
DI2
A[6:0]
WA[6:0]
CLK
WE
RAM
AI
AX
A input
6
DI1
DI2
A[6:0]
WA[6:0]
CLK
WE
UG474_c5_04_110110
Figure 5-4:
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Simplified 7 Series FPGA SLICEM Distributed RAM
www.xilinx.com
Send Feedback
63
Chapter 5: Timing
Distributed RAM Timing Parameters
Table 5-4 shows the timing parameters for the distributed RAM in SLICEM for a majority
of the paths in Figure 5-4.
Table 5-4:
Distributed RAM Timing Parameters
Parameter
Function
Description
Sequential Delays for a Slice LUT Configured as RAM (Distributed RAM)
TSHCKO
CLK to A/B/C/D outputs
Time after the CLK of a write operation that the
data written to the distributed RAM is stable on
the A/B/C/D outputs of the slice.
TSHCKO (TSHCKO_1)
CLK to AMUX/BMUX/CMUX/DMUX
outputs
Time after the CLK of a write operation that the
data written to the distributed RAM is stable on
the AMUX/BMUX/CMUX/DMUX outputs of
the slice.
Setup and Hold Times for a Slice LUT Configured as RAM (Distributed RAM)(1)
AI/BI/CI/DI configured as data inputs
(DI1)(2)
Time before/after the clock that data must be
stable at the AI/BI/CI/DI inputs of the slice
(configured as RAM).
TAS/TAH
(TAS_LRAM/TAH_LRAM)
A/B/C/D address inputs
Time before/after the clock that address signals
must be stable at the A/B/C/D LUT inputs of
the slice (configured as RAM).
TWS/TWH
(TWS_LRAM/TWH_LRAM)
WE input
Time before/after the clock that the write
enable signal must be stable at the WE input of
the slice (configured as RAM).
TDS/TDH
(TDS_LRAM/TDH_LRAM)
Clock CLK
TMPW
(TMPW_LRAM)
Minimum clock pulse width for distributed
RAM.
TMCP
Minimum clock period to meet address write
cycle time.
Notes:
1. TXS = Setup Time (before clock edge), and TXH = Hold Time (after clock edge).
2. Parameter includes AX/BX/CX/DX configured as a data input (DI2).
64
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Slice Distributed RAM Timing Model and Parameters (Available in SLICEM Only)
Distributed RAM Timing Characteristics
The timing characteristics of a distributed RAM implemented in a 7 series FPGA slice
(LUT configured as RAM) are shown in Figure 5-5.
X-Ref Target - Figure 5-5
1
2
3
4
5
6
7
TMCP
TMPW
TMPW
CLK
TAS
A/B/C/D
(ADDR)
2
F
3
4
0
1
5
E
TDS
AI/BI/CI/DI
(DI)
1
X
TWS
0
X
TILO
TILO
WE
DATA_OUT
A/B/C/D
Output
TSHCKO
1
WRITE
MEM(F)
READ
0
1
WRITE
WRITE
0
WRITE
MEM(E)
READ
UG474_c5_05_110110
Figure 5-5: Slice Distributed RAM Timing Characteristics
Clock Event 1: Write Operation
During a Write operation, the contents of the memory at the address on the ADDR inputs
are changed. The data written to this memory location is reflected on the A/B/C/D
outputs synchronously.
•
At time TWS before clock event 1, the write-enable signal (WE) becomes valid-High,
enabling the RAM for a Write operation.
•
At time TAS before clock event 1, the address (2) becomes valid at the A/B/C/D
inputs of the RAM.
•
At time TDS before clock event 1, the DATA becomes valid (1) at the DI input of the
RAM and is reflected on the A/B/C/D outputs at time TSHCKO after clock event 1.
This is also applicable to the AMUX, BMUX, CMUX, DMUX, and COUT outputs at
time TSHCKO after clock event 1.
Clock Event 2: Read Operation
All Read operations are asynchronous in distributed RAM. As long as WE is Low, the
address bus can be asserted at any time. The contents of the RAM on the address bus are
reflected on the A/B/C/D outputs after a delay of length TILO (propagation delay through
a LUT). The address (F) is asserted after clock event 2, and the contents of the RAM at
address (F) are reflected at the output after a delay of length TILO.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
65
Chapter 5: Timing
CLB Slice SRL Shift Register Timing Model and Parameters
(Available in SLICEM Only)
Figure 5-6 illustrates shift register implementation in a 7 series FPGA slice. Some elements
of the slice have been omitted for clarity. Only the elements relevant to the timing paths
described in this section are shown.
X-Ref Target - Figure 5-6
SRL
DI1
DI
O6
D address
6
A
D
MC31
CLK
WE
CLK
W
SRL
DI1
O6
CI
C address
6
A
C
MC31
CLK
WE
SRL
DI1
O6
BI
B address
6
A
B
MC31
CLK
WE
SRL
DI1
O6
AI
A address
6
A
CLK
MC31
A
DMUX
WE
UG474_c5_06_110110
Figure 5-6:
66
Send Feedback
Simplified 7 Series FPGA Slice SRL
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Slice SRL Shift Register Timing Model and Parameters (Available in SLICEM Only)
Slice SRL Timing Parameters
Table 5-5 shows the SLICEM SRL timing parameters for a majority of the paths in
Figure 5-6.
Table 5-5:
Slice SRL Timing Parameters
Parameter
Function
Description
Sequential Delays for a Slice LUT Configured as an SRL
TREG(1)
CLK to A/B/C/D outputs
Time after the CLK of a write operation that the
data written to the SRL is stable on the A/B/C/
D outputs of the slice.
TREG (TREG_MUX)(1)
CLK to AMUX - DMUX outputs
Time after the CLK of a write operation that the
data written to the SRL is stable on the AMUX/
BMUX/CMUX/DMUX outputs of the slice.
TREG (TREG_M31)
CLK to DMUX output via MC31
output
Time after the CLK of a write operation that the
data written to the SRL is stable on the DMUX
output via MC31 output.
Setup and Hold Times for a Slice LUT Configured as an SRL(2)
TCECK/TCKCE
(TCECK_SHFREG/TCKCESHFREG)
CE input (CE)
Time before/after the clock that the
clock-enable signal must be stable at the CE
input of the slice LUT (configured as an SRL).
TWS/TWH
(TWS_SHFREG/TWH_SHFREG)
CE input (WE)
Time before/after the clock that the
write-enable signal must be stable at the WE
input of the slice LUT (configured as an SRL).
TDS/TDH(3)
(TDS_SHFREG/TDH_SHFREG)
AI/BI/CI/DI configured as data
inputs (DI)
Time before/after the clock that the data must
be stable at the AI/BI/CI/DI inputs of the slice
(configured as an SRL).
Clock CLK
Minimum clock pulse width for distributed
RAM.
TMPW
Notes:
1. This parameter includes a LUT configured as a two-bit shift register.
2. TXS = Setup Time (before clock edge), and TXH = Hold Time (after clock edge).
3. Parameter includes AX/BX/CX/DX configured as a data input (DI2) or two bits with a common shift.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
67
Chapter 5: Timing
Slice SRL Timing Characteristics
Figure 5-7 illustrates the timing characteristics of a shift register implemented in a 7 series
FPGA slice (a LUT configured as an SRL).
X-Ref Target - Figure 5-7
1
2
3
4
5
6
32
CLK
TWS
Write Enable
(WE)
TDS
Shift_In (DI)
0
Address
(A/B/C/D)
1
1
0
0
TREG
1
0
2
1
TILO
Data Out
(A/B/C/D)
X
1
MSB
(MC31/DMUX)
0
TREG
X
X
X
1
0
X
TILO
1
1
X
0
X
1
X
0
UG474_c5_07_110110
Figure 5-7:
Slice SRL Timing Characteristics
Clock Event 1: Shift In
During a write (Shift In) operation, the single-bit content of the register at the address on
the A/B/C/D inputs is changed, as data is shifted through the SRL. The data written to
this register is reflected on the A/B/C/D outputs synchronously, if the address is
unchanged during the clock event. If the A/B/C/D inputs are changed during a clock
event, the value of the data at the addressable output (A/B/C/D outputs) is invalid.
•
At time TWS before clock event 1, the write-enable signal (WE) becomes valid-High,
enabling the SRL for the Write operation that follows.
•
At time TDS before clock event 1, the data becomes valid (0) at the DI input of the SRL
and is reflected on the A/B/C/D output after a delay of length TREG after clock
event 1. Because address 0 is specified at clock event 1, the data on the DI input is
reflected at A/B/C/D output, because it is written to register 0.
Clock Event 2: Shift In
•
At time TDS before clock event 2, the data becomes valid (1) at the DI input of the SRL
and is reflected on the A/B/C/D output after a delay of length TREG after clock
event 2. Because address 0 is still specified at clock event 2, the data on the DI input is
reflected at the D output, because it is written to register 0.
Clock Event 3: Shift In/Addressable (Asynchronous) Read
All Read operations are asynchronous to the CLK signal. If the address is changed
(between clock events), the contents of the register at that address are reflected at the
addressable output (A/B/C/D outputs) after a delay of length TILO (propagation delay
through a LUT).
•
68
At time TDS before clock event 3, the data becomes valid (1) at the DI input of the SRL
and is reflected on the A/B/C/D output TREG time after clock event 3.
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
CLB Slice SRL Shift Register Timing Model and Parameters (Available in SLICEM Only)
•
The address is changed (from 0 to 2). The value stored in register 2 at this time is a 0
(in this example, this was the first data shifted in), and it is reflected on the A/B/C/D
output after a delay of length TILO.
Clock Event 32: MSB (Most Significant Bit) Changes
•
At time TREG after clock event 32, the first bit shifted into the SRL becomes valid
(logical 0 in this case) on the DMUX output of the slice via the MC31 output of LUT A
(SRL). This is also applicable to the AMUX, BMUX, CMUX, DMUX, and COUT
outputs at time TREG after clock event 1.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
69
Chapter 5: Timing
70
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Chapter 6
Advanced Topics
Using the Latch Function as Logic
Because the latch function is level-sensitive, it can be used as the equivalent of a logic gate.
The primitives to specify this function are AND2B1L (a 2-input AND gate with one input
inverted) and OR2L (a 2-input OR gate), as shown in Figure 6-1.
X-Ref Target - Figure 6-1
AND2B1L
OR2L
UG474_c6_01_110210
Figure 6-1:
AND2B1L and OR2L Components
As shown in Figure 6-2, the data and SR inputs and Q output of the latch are used when the
AND2B1L and OR2L primitives are instantiated, and the CK gate and CE gate enables are
held active-High. The AND2B1L combines the latch data input (the inverted input on the
gate, DI) with the asynchronous clear input (SRI). The OR2L combines the latch data input
with an asynchronous preset. Generally, the latch data input comes from the output of a
LUT within the same slice, extending the logic capability to another external input.
Because there is only one SR input per slice, using more than one AND2B1L or OR2L per
slice requires a shared common external input.
X-Ref Target - Figure 6-2
IN[6:1]
LUT6
O6
D
Q
Q
VCC
CE
CK
SR
SRI
UG474_c6_02_110310
Figure 6-2: Implementation of OR2L (Q = D or SRI)
The device model shows these functions as AND2L and OR2L configurations of the
storage element.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
71
Chapter 6: Advanced Topics
The AND2B1L and OR2L two-input gates save LUT resources and are initialized to a
known state on power-up and on GSR assertion (Global Set/Reset). Using these primitives
can reduce logic levels and increase logic density of the device by trading register/latch
resources for logic. However, due to the static inputs required on the clock and clock
enable inputs, specifying one or more AND2B1L or OR2L primitives can cause register
packing and density issues in a slice disallowing the use of the remaining registers and
latches.
Interconnect Resources
Interconnect is the programmable network of signal pathways between the inputs and
outputs of functional elements within the FPGA, such as IOBs, CLBs, DSP slices, and block
RAM. Interconnect, also called routing, is segmented for optimal connectivity. The Xilinx
placement and routing tools exploit the rich interconnect array to deliver optimal system
performance and the fastest compile times.
The 7 series CLBs are arranged in a regular array inside the FPGA. Each connects to a
switch matrix for access to the general-routing resources, which run vertically and
horizontally between the CLB rows and columns. A similar switch matrix connects other
resources, such as the DSP slices and block RAM resources.
Most of the interconnect features are transparent to FPGA designers. Knowledge of the
interconnect details can be used to guide design techniques but is not necessary for
efficient FPGA design. Only selected types of interconnect are under user control. These
include the clock routing resources, which are selected by using clock buffers and
discussed in more detail in UG472, 7 Series FPGAs Clocking Resources User Guide. Two
global control signals, GSR and GTS, are selected by using the STARTUPE2 primitive.
Global Controls GSR and GTS
In addition to the general-purpose interconnect, 7 series FPGAs have two global logic
control signals, as described in Table 6-1.
Table 6-1:
Global Logic Control Signals
Global Control Input
Description
GSR
Global Set/Reset: When High, asynchronously places all registers
and flip-flops in their initial state.
GTS
Global 3-State: When High, asynchronously forces all I/O pins to a
high-impedance state (High-Z, 3-state).
If a common initialization signal is needed for every flip-flop in the design, use the GSR
control in a design instead of a separate routed global reset signal to make CLB inputs
available, which results in a smaller more efficient design. The GSR signal must always
re-initialize every flip-flop. Using GSR and GTS does not use any general-purpose routing
resources. The GSR signal is asserted automatically during the FPGA configuration
process, guaranteeing that the FPGA starts up in a known state. At configuration, because
flip-flops are enabled by the configuration clock and then clocked by a user clock, it is
recommended to enable key flip-flops or key clock signals on the user clock after
configuration.
72
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
Devices Using Stacked Silicon Interconnect (SSI) Technology
STARTUPE2 Primitive
The GSR and GTS signal sources are defined and connected using the STARTUPE2
primitive. This primitive allows the user to define the source of these dedicated nets. GSR
and GTS are always active during configuration, and connecting signals to them on the
STARTUPE2 primitive defines how they are controlled after configuration. By default,
they are disabled after configuration on a selected clock cycle of the start-up phase,
enabling the flip-flops and I/Os in the device. The STARTUPE2 primitive also includes
other signals used specifically during configuration. For more information, refer to the
UG470, 7 Series FPGAs Configuration User Guide.
Interconnect Optimization
Interconnect delays vary according to the specific implementation and loading in a design.
The type of interconnect, distance required to travel in the device, and number of switch
matrices to traverse factor into the total delay. Most timing issues are addressed by
examining the block delays and determining the impact of using fewer levels or faster
paths. If interconnect delays seem too long, increase implementation effort levels or
iterations to improve performance along with making sure that the required timing is in
the constraints file.
Nets with critical timing or that are heavily loaded can often be improved by replicating
the source of the net. The dual 5-input LUT configuration of the slice simplifies the
replication of logic in the same slice, which minimizes any additional loads on the inputs to
the source function. Replicating logic in multiple slices gives the software more flexibility
to place the sources independently.
Floorplanning is the process of specifying user-placement constraints. Floorplanning can
be done either before or after automatic place and route, but automatic place and route is
always recommended first before specifying user floorplanning. The Xilinx
implementation tool provides a graphical view of placement. It helps the designer choose
between RTL coding and synthesis and implementation, with extensive design exploration
and analysis features.
Devices Using Stacked Silicon Interconnect (SSI) Technology
As noted in the DS180, 7 Series FPGAs Overview, some of the Virtex-7 devices use stacked
silicon interconnect (SSI) technology. These devices provide a unique additional type of
interconnect resource called a super long line or SLL. These special routing resources can
be treated like other interconnect resources as they are abundant (more than 10,000) and
fast (about 1 ns). The SLL provides a connection between super logic regions (SLRs), as
shown in Figure 6-3. Combining multiple SLRs effectively increases the height of the
ASMBL™ architecture columns and increases the overall capacity of the device. The
software can be allowed to take best advantage of this configuration, or users can apply
floorplanning to control placement within and between SLRs. Note that carry logic
cascading is limited to within an SLR and does not continue through the SLL connections.
The same is true of other types of cascading including block RAM, DSP, and DCI.
Boundaries are visible in floorplanning tools and reported in timing reports. The number
of SLRs per Virtex-7 device is listed in the 7 Series FPGAs Overview.
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
www.xilinx.com
Send Feedback
73
Chapter 6: Advanced Topics
X-Ref Target - Figure 6-3
Super Logic
Region (SLR)
Super Long
Line (SLL)
Super Logic
Region (SLR)
UG474_c6_03_090211
Figure 6-3:
Stacked Silicon Interconnect Technology
Designs targeting devices using SSI technology only require one STARTUPE2
instantiation. One source for GSR and GTS signals propagates across the SLR boundaries
to all elements of the device.
Optimization for devices using SSI technology can start with the same techniques as for
any standard device, including using proper design techniques, timing constraints or
options to automatically find the best implementation. Floorplanning for these devices can
use the same graphical tools.
For more information on devices using SSI technology, see the following:
74
•
DS180, 7 Series FPGAs Overview
•
Stacked Silicon Interconnect Technology page
www.xilinx.com/technology/roadmap/stacked-silicon-interconnect
•
WP380, Xilinx Stacked Silicon Interconnect Technology Delivers Breakthrough FPGA
Capacity, Bandwidth, and Power Efficiency
Send Feedback
www.xilinx.com
7 Series FPGAs CLB User Guide
UG474 (v1.7) November 17, 2014
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

advertisement