Title: Device and Architecture Co-Optimization for FPGA Power Reduction
1Device and Architecture Co-Optimization for FPGA
Power Reduction
- Lerong Cheng, Phoebe Wong,
- Fei Li, Yan Lin,
- and Prof. Lei He
- EE Department, UCLA
- Partially supported by NSF CAREER award
CCR-0093273/0401682 and NSF grant CCR-0306682.
Address comments to lhe_at_ee.ucla.edu
2Outline
- Background and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Conclusion
3Evaluation of Conventional FPGA Architecture
- LUT size and cluster size have been evaluated for
conventional FPGA - performance and area Ahmed et al, ISFPGA00
- power and performance Li et al, ISFPGA 03
- Architecture tuning leads to 2.8X energy
difference and1.5X delay difference
Island style FPGA architecture
Evaluation result
4Evaluation of Low-Power FPGA Architecture
- Field programmable dual-vdd for power reduction
Lin et al, ISFPGA05 - Applying field programmable dual Vdd reduces
energy-delay product by 49
5Evaluation Methodology
Benchmark circuits
Logic Optimization(SIS)
Tech-Mapping (RASP)
Timing-Driven Packing (TV-Pack)
Cycle-accurate Power Simulator (Psim)
Placement Routing (VPR)
Area
Delay
6Impact of Device Tuning
- All the previous work only considers architecture
tuning - Device tuning leads to 84X power difference and
12X delay difference - It is necessary to perform device tuning and
architecture tuning simultaneously
7Challenge of Device and Architecture
Co-Optimization
- We consider the following architecture and device
parameters during our co-optimization - Architecture parameters
- Cluster size (N)
- LUT size (K)
- Device parameters
- Supply voltage (Vdd)
- Threshold voltage (Vt)
- Hyper-architecture (hyper-arch) is the
combination of the device and architecture
parameters. - Large number of hyper-arch combinations
- VPR and Psim are too slow to deal with such large
numberof experiments - Need fast yet accurate power and delay estimation
8Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Trace collection
- Trace based power and delay model
- Accuracy and efficiency verification of
Trace-based estimator - Device and architecture co-optimization
- Conclusion
9Trace Collection
Circuit element statistics
Critical path structure
VPR and Psim
Short circuit power ratio
Trace
Ptrace
Switching activity
Area
- Assume trace information will remain the same
when device setting changes
10Trace Base Estimation (Ptrace) Framework
Device independent
Ptrace
Trace
Chip level delay, power, and area
Device dependent
Circuit level delay and power
11Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Trace collection
- Trace based power and delay model
- Accuracy and efficiency verification of
Trace-based estimator - Device and architecture co-optimization
- Conclusion
12Delay Model in VPR
- Delay is calculated for each path as
- Nip is number of type i elements in the path and
Di is delay oftype i element - Delay of the logic elements is measured by SPICE
simulation - Elmore delay is used for interconnect wire
segments - Critical path is the path with longest delay
13Delay in Ptrace
- Obtain the path structure of a set of longest
circuit paths - Assume that when device setting changes, the new
critical path is still among the set of longest
paths. - Delay computation
Trace information
Device dependent parameters
14Dynamic Power Model
- Psim
- Switch power
- Switching activity is measured by timing
simulation for each node - Si is the average switching activity
- Short circuit power
- asc is calculated for each node
- Ptrace
- Switch power
- Short circuit power
- asc is the average short circuit power ratio for
the whole circuit
Trace information
Device dependent parameters
15Static Power Model
- Psim
- Without power gating
- With power gating
- Ptrace
- Without power gating
- With power gating
Trace information
Device dependent parameters
16Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Trace collection
- Trace based power and delay model
- Accuracy and efficiency verification of
Trace-based estimator - Device and architecture co-optimization
- Conclusion
17Experiment Setting
- Collect trace using ITRS 70nm technology, but
apply to both 100nm and 70nm technologies - 20 MCNC benchmarks
- Assume each benchmark works in its highest
possible frequency - Power and delay are computed as geometric mean
of20 benchmarks. - Evaluation range
Vdd Vt LUT size (K) Cluster size (N)
0.81.1 0.20.4 37 612
18Accuracy
- Average power error is 3.4.
- Average delay error is 6.4.
- Delay error is due to Ptrace ignores the impact
of path branches that considered in VPR
19Runtime
- VPR and Psim for one device setting
- five days on eight 1.2GHz Intel Xeon servers
- Ptrace for 20 device settings
- 80 seconds on one 1.2GHz Intel Xeon server
20Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Energy and delay tradeoff
- ED and area tradeoff
- Comparison between classes
- Comparison between device tuning and architecture
tuning - Conclusion
21Architectures Classes to be Evaluated
- Hyper-architecture classes
-
- Baseline case
- Vdd suggested by ITRS
- Architecture same as Xilinx Virtex-II.
- Vt optimized by our method with respect to the
above architecture and Vdd
Hyper-arch classes Vt
Homo-Vt Homogeneous Vt
Hetero-Vt Heterogeneous Vt
Homo-VtG Homogeneous Vt Power Gating
Hetero-VtG Heterogeneous Vt Power Gating
Vdd Vt LUT size (K) Cluster size (N)
0.9 0.3 4 8
22Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Energy and delay tradeoff
- ED and area tradeoff
- Comparison between classes
- Comparison between device tuning and architecture
tuning - Conclusion
23Energy and Delay Tradeoff
- Dominant hyper-arch
- Hyper-arch B is inferior to A if A has less
energy and smaller delay than B. - Dominant hyper-archs (dom-arch) are the
hyper-archs that are NOT inferior to any other
hyper-archs.
24Energy and Delay Tradeoff
- Hetero-Vt can reduce power
- Power gating reduces more leakage power than
hetero-Vt - Hetero-Vt has less impact when power gating is
applied
25Min-ED Hyper-Arch
Hyper-arch classes Vdd (V) CVt (V) IVt (V) (N, K) ED (nJns) ED reduction
Baseline 0.9 0.3 0.3 (8,4) 26.9 -
Homo-Vt 0.9 0.3 0.3 (6,7) 23.3 13.4
Hetero-Vt 0.9 0.2 0.25 (8,4) 21.4 20.5
Homo-VtG 0.9 0.25 0.25 (12,4) 11.1 58.9
Hetero-VtG 0.9 0.2 0.25 (8,4) 11 59.0
- To achieve the best energy and delay tradeoff, we
find out the hyper-arch with the minimum energy
and delay product (ED) - Compared to the baseline, the min-ED hyper-arch
of the conventional FPGA (Homo-Vt) reduces ED by
13.4 - For the Hetero-Vt class, ED is reduced by 20.5
- If power gating is applied, ED can be reduced by
up to 59.0
26Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Energy and delay tradeoff
- ED and area tradeoff
- Comparison between classes
- Comparison between device tuning and architecture
tuning - Conclusion
27ED and area Tradeoff
- Architecture tuning has great impact on area.
- To achieve the best area and ED tradeoff, we find
the hyper-arch with the minimum product of area,
energy and delay (AED)
28ED Area Tradeoff for Classes without Power Gating
- Compared to the min-ED hyper arch, the min-AED
hyper-arch significantly reduce area with a small
ED increase
29Sleep Transistor Size Tuning
- When Power gating is applied, sleep transistors
may increase area - The larger the sleep transistor size, the smaller
the delay - Sleep transistor size tuning
- Area overhead introduced by sleep transistors of
logic blocks is negligible. - We consider 2X, 4X, 7X and 10X PMOS as sleep
transistor for switch buffer
30ED Area Tradeoff for Classes with Power Gating
- The area reduction achieved by device and
architecture co-optimization compensates the
area overhead introduced by sleep transistors
31Min-AED Hyper-Arch
Vdd (V) CVt (V) IVt (V) (N,K) Sleep transistor size ED (nJns) Normalized area AED reduction
Baseline 0.9 0.30 0.30 (8,4) - 26.9 1.00 -
Homo-Vt 1.0 0.30 0.30 (6,4) - 23.6 0.80 30.0
Hetero-Vt 0.9 0.30 0.25 (12, 4) - 21.3 0.77 40.0
Hetero-VtG 0.9 0.25 0.25 (12, 4) 2 12.4 0.92 57.6
Hetero-VtG 0.9 0.20 0.25 (12, 4) 2 12.2 0.92 58.3
- Compared to the baseline, the min-AED hyper-arch
in the conventional FPGA class can reduce area by
20 and ED by 12.3 - In the Hetero-Vt class, ED is reduced by 20.8
and area is reduced by 23 compared to the
baseline - If power gating is applied, ED is reduced by
54.6 and area is reduced by 8.3
32Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Energy and delay tradeoff
- ED and area tradeoff
- Comparison between classes
- Comparison between device tuning and architecture
tuning - Conclusion
33Comparison Between Classes in Similar Performance
Range
Homo-Vt Homo-Vt Homo-Vt Homo-Vt Homo-Vt Homo-Vt Hetero-Vt Hetero-Vt Hetero-Vt Hetero-Vt Hetero-Vt Hetero-Vt Hetero-Vt
Vdd Vt (N, K) E (nJ) D (ns) ED (nJns) Vdd CVt IVt (N, K) E (nJ) D (ns) ED (nJns)
0.9 0.30 6,6 1.33 18.6 24.8 0.9 0.3 0.35 6,4 1.16 20.1 23.3
0.9 0.3 10,5 1.27 19.8 25 0.9 0.3 0.35 12,4 1.14 20.5 23.7
0.9 0.3 6,4 1.23 21.6 26.5 0.9 0.3 0.35 8,4 1.09 22.1 24.1
Homo-VtG Homo-VtG Homo-VtG Homo-VtG Homo-VtG Homo-VtG Hetero-VtG Hetero-VtG Hetero-VtG Hetero-VtG Hetero-VtG Hetero-VtG Hetero-VtG
Vdd Vt (N, K) E (nJ) D (ns) ED (nJns) Vdd CVt IVt (N, K) E (nJ) D (ns) ED (nJns)
0.8 0.25 10,5 0.70 19.4 13.7 0.9 0.25 0.3 12,4 0.66 18.9 12.5
0.8 0.25 8,4 0.62 20.9 12.9 0.8 0.25 0.25 8,4 0.62 20.9 12.9
0.8 0.25 12,4 0.62 21 12.9 0.8 0.25 0.25 12,4 0.62 21 12.9
- Vt for logic block is lower than Vt for
interconnect - Vt for classes with power gating is lower
34Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Energy and delay tradeoff
- ED and area tradeoff
- Comparison between classes
- Comparison between device tuning and architecture
tuning - Conclusion
35Dom-Archs under Different Device Settings
- For a given device setting architecture tuning
changes delay and energy in a smaller range - Device tuning has a much more impact on delay and
energy
36Outline
- Back ground and motivation
- Trace-based power and delay estimation
- Device and architecture co-optimization
- Conclusion
37Conclusion and Discussion
- Trace-based estimator provides efficient and
accurate FPGA power and delay estimation - Average power error is 3.4 and average delay
error is 6.1 - Device and architecture co-optimization reduces
ED by 20.5 and area by 23.3 when there is no
power gating - With power gating, device and architecture
co-optimization reduces ED by 54.6 and area by
8.3 - Device tuning has a more significant impact on
delay and power than architecture tuning does - In recent research, Ptrace has been extended to
consider leakage and timing yield with process
variations
38Thank you