Cited By
View all- Barkalov ATitarenko LKrzywicki K(2022)Improving Hardware in LUT-Based Mealy FSMsApplied Sciences10.3390/app1216806512:16(8065)Online publication date: 11-Aug-2022
postplacement | postrouting | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
lower bound | covered | |||||||||||
circuit | ref. | \(W = 0\) | \(W = 1\) | \(W = 0\) | \(W = 1\) | legal. | \(-\Delta [\%]\) | ref. | w/ dir. | mux | \(-\Delta [\%]\) | w/o dir. |
raygentop | 4.70 | 4.70 | 4.70 | 4.70 | 4.70 | 4.70 | 0.00 | 4.87 | 4.88 | 0.02 | \(-\) 0.21 | 4.88 |
ch_intrinsics | 3.15 | 3.15 | 2.84 | 3.15 | 3.15 | 3.15 | 0.00 | 3.28 | 3.27 | 0.03 | 0.30 | 3.27 |
mkDelayWorker32B | 6.83 | 6.83 | 6.55 | 6.83 | 6.58 | 6.58 | 3.66 | 7.09 | 7.04 | 0.03 | 0.71 | 7.36 |
mkSMAdapter4B | 5.16 | 5.11 | 4.97 | 5.11 | 5.02 | 5.02 | 2.71 | 5.38 | 5.26 | 0.05 | 2.23 | 5.63 |
bgm | 23.56 | 23.56 | 22.21 | 23.56 | 22.66 | 22.66 | 3.82 | 23.66 | 23.04 | 0.20 | 2.62 | 26.33 |
boundtop | 6.10 | 6.01 | 5.57 | 6.01 | 5.73 | 5.73 | 6.07 | 6.05 | 5.82 | 0.05 | 3.80 | 6.37 |
stereovision0 | 3.74 | 3.74 | 3.31 | 3.74 | 3.52 | 3.52 | 5.88 | 3.74 | 3.57 | 0.06 | 4.55 | 4.06 |
diffeq1 | 20.45 | 19.48 | 18.24 | 19.81 | 19.19 | 19.19 | 6.16 | 21.16 | 20.01 | 0.12 | 5.43 | 21.86 |
diffeq2 | 15.69 | 14.92 | 13.46 | 15.02 | 14.48 | 14.48 | 7.71 | 16.14 | 15.14 | 0.11 | 6.20 | 16.68 |
blob_merge | 9.90 | 8.76 | 6.79 | 9.44 | 8.90 | 9.16 | 7.47 | 9.89 | 9.21 | 0.11 | 6.88 | 10.56 |
or1200 | 13.08 | 12.66 | 10.76 | 12.77 | 11.69 | 11.75 | 10.17 | 13.12 | 12.20 | 0.23 | 7.01 | 15.66 |
LU8PEEng | 105.05 | 101.07 | 91.47 | 101.49 | 95.57 | 95.63 | 8.97 | 104.86 | 96.45 | 0.98 | 8.02 | 110.17 |
sha | 11.89 | 11.02 | 9.15 | 11.25 | 10.65 | 10.83 | 8.92 | 11.88 | 10.86 | 0.15 | 8.59 | 12.59 |
geomean | 10.29 | 9.99 | 9.03 | 10.09 | 9.68 | 9.72 | 5.54 | 10.46 | 10.01 | 0.09 | 4.30 | 11.07 |
size | components | degrees | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
circuit | \(|V|\) | \(|E|\) | W | H | # | \(\langle |V|\rangle\) | \(\text{max} |V|\) | \(\langle \text{total}\rangle\) | max total | max in | max out | diameter |
boundtop | 29 | 18 | 18 | 9 | 11 | 2.64 | 5 | 1.24 | 3 | 2 | 1 | 4 |
25 | 14 | 18 | 9 | 11 | 2.27 | 3 | 1.12 | 2 | 2 | 1 | 3 | |
stereovision0 | 51 | 39 | 39 | 13 | 12 | 4.25 | 16 | 1.53 | 5 | 4 | 2 | 6 |
36 | 20 | 39 | 12 | 16 | 2.25 | 3 | 1.11 | 2 | 2 | 2 | 3 | |
diffeq1 | 38 | 32 | 7 | 6 | 6 | 6.33 | 14 | 1.68 | 6 | 1 | 5 | 9 |
31 | 19 | 7 | 6 | 12 | 2.58 | 4 | 1.23 | 3 | 1 | 2 | 4 | |
diffeq2 | 38 | 33 | 6 | 4 | 5 | 7.60 | 17 | 1.74 | 4 | 3 | 2 | 11 |
30 | 17 | 6 | 4 | 13 | 2.31 | 3 | 1.13 | 2 | 2 | 1 | 3 | |
blob_merge | 49 | 37 | 4 | 8 | 15 | 3.27 | 9 | 1.51 | 5 | 2 | 4 | 6 |
47 | 28 | 4 | 8 | 19 | 2.47 | 3 | 1.19 | 2 | 2 | 2 | 3 | |
or1200 | 97 | 71 | 14 | 16 | 26 | 3.73 | 11 | 1.46 | 3 | 3 | 2 | 11 |
87 | 50 | 14 | 15 | 37 | 2.35 | 4 | 1.15 | 3 | 3 | 1 | 3 | |
LU8PEEng | 334 | 227 | 24 | 21 | 111 | 3.01 | 10 | 1.36 | 5 | 4 | 5 | 8 |
294 | 164 | 24 | 21 | 130 | 2.26 | 4 | 1.12 | 3 | 3 | 2 | 3 | |
sha | 74 | 52 | 13 | 8 | 23 | 3.22 | 10 | 1.41 | 4 | 2 | 3 | 7 |
59 | 33 | 13 | 8 | 26 | 2.27 | 4 | 1.12 | 2 | 2 | 2 | 4 |
ILP preparation | ILP status | flexibility (last feas.) | \(\langle \text{ILP sol. t}\rangle\) | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
circuit | #LUTs | wall clock | CPU | \(\langle \text{LP sol. t}\rangle\) | \(\langle \text{setup t}\rangle\) | feas. | infeas. | timeout | \(\langle \text{pos./u}\rangle\) | \(\langle \text{pair/e}\rangle\) | \(\langle |\text{tree}|\rangle\) | feas. | infeas. |
diffeq2 | 322 | 92.05 | 370.63 | 0.05 | 3.12 | 4 | 0 | 0 | 42 | 81 | 73 | 12.88 | — |
diffeq1 | 485 | 156.86 | 810.27 | 0.07 | 3.69 | 5 | 0 | 0 | 43 | 78 | 215 | 22.93 | — |
mkSMAdapter4B | 1982 | 52.82 | 34.92 | 0.33 | 2.38 | 2 | 1 | 0 | 32 | 37 | 0 | 0.06 | 0.04 |
sha | 2280 | 2355.68 | 38358.80 | 0.21 | 6.44 | 5 | 0 | 2 | 38 | 88 | 4561 | 210.42 | — |
or1200 | 3054 | 279.96 | 1423.68 | 0.11 | 6.71 | 3 | 0 | 0 | 44 | 85 | 682 | 59.12 | — |
boundtop | 3070 | 87.45 | 73.94 | 0.22 | 3.64 | 4 | 1 | 0 | 31 | 50 | 0 | 0.44 | 0.02 |
mkDelayWorker32B | 5602 | 127.45 | 110.88 | 0.26 | 5.44 | 3 | 0 | 0 | 42 | 39 | 0 | 0.09 | — |
blob_merge | 6019 | 2967.47 | 47072.96 | 0.46 | 4.76 | 5 | 0 | 2 | 42 | 93 | 9455 | 311.82 | — |
stereovision0 | 14779 | 274.80 | 271.00 | 0.41 | 10.13 | 3 | 1 | 0 | 36 | 79 | 0 | 2.13 | 0.02 |
LU8PEEng | 26455 | 4637.07 | 55084.80 | 3.43 | 72.80 | 3 | 0 | 4 | 37 | 81 | 4666 | 264.04 | — |
bgm | 36480 | 1624.23 | 3445.75 | 3.38 | 71.54 | 4 | 0 | 0 | 36 | 82 | 5412 | 43.03 | — |
postplacement | postrouting | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
circuit | ref. | \(W = 1\) | \(-\Delta\) [%] | \(W = 2\) | \(-\Delta\) [%] | ref. | \(W = 1\) | \(-\Delta\) [%] | \(W = 2\) | \(-\Delta\) [%] |
raygentop | 4.70 | 4.70 | 0.00 | 4.70 | 0.00 | 4.87 | 4.88 | \(-\) 0.21 | 4.88 | \(-\) 0.21 |
ch_intrinsics | 3.15 | 3.15 | 0.00 | 3.15 | 0.00 | 3.28 | 3.30 | \(-\) 0.61 | 3.29 | \(-\) 0.30 |
mkDelayWorker32B | 6.83 | 6.63 | 2.93 | 6.58 | 3.66 | 7.09 | 7.06 | 0.42 | 7.04 | 0.71 |
mkSMAdapter4B | 5.16 | 5.02 | 2.71 | 5.02 | 2.71 | 5.38 | 5.31 | 1.30 | 5.25 | 2.42 |
bgm | 23.56 | 22.85 | 3.01 | 22.66 | 3.82 | 23.66 | 23.20 | 1.94 | 22.84 | 3.47 |
boundtop | 6.10 | 5.89 | 3.44 | 5.69 | 6.72 | 6.05 | 5.91 | 2.31 | 5.73 | 5.29 |
stereovision0 | 3.74 | 3.74 | 0.00 | 3.74 | 0.00 | 3.74 | 3.76 | \(-\) 0.53 | 3.76 | \(-\) 0.53 |
diffeq1 | 20.45 | 19.62 | 4.06 | 19.42 | 5.04 | 21.16 | 20.33 | 3.92 | 20.20 | 4.54 |
diffeq2 | 15.69 | 14.97 | 4.59 | 14.61 | 6.88 | 16.14 | 15.46 | 4.21 | 15.18 | 5.95 |
blob_merge | 9.90 | 9.26 | 6.46 | 9.17 | 7.37 | 9.89 | 9.26 | 6.37 | 9.23 | 6.67 |
or1200 | 13.08 | 12.13 | 7.26 | 11.95 | 8.64 | 13.12 | 12.50 | 4.73 | 12.21 | 6.94 |
LU8PEEng | 105.05 | 97.51 | 7.18 | 94.55 | 10.00 | 104.86 | 98.03 | 6.51 | 95.03 | 9.37 |
sha | 11.89 | 10.97 | 7.74 | 10.87 | 8.58 | 11.88 | 11.01 | 7.32 | 10.92 | 8.08 |
geomean | 10.29 | 9.90 | 3.79 | 9.78 | 4.96 | 10.46 | 10.16 | 2.87 | 10.04 | 4.02 |
FPGA application developers often use pipelining, C-slowing and retiming to improve the performance of their designs. Unfortunately, registered netlists present a fundamentally different problem to CAD tools, potentially limiting the benefit of these ...
Modern FPGA device contains complex clocking architecture on top of FPGA logic fabric. To best utilize FPGA clocking architecture, both FPGA designers and EDA tool developers need to understand the clocking architecture and design best methodology/...
Placement is one of the most important steps in the RTL-to-GDSII synthesis process as it directly defines the interconnects. The rapid increase in IC design complexity and the widespread use of intellectual-property blocks have made the so-called mixed-...
Association for Computing Machinery
New York, NY, United States
View or Download as a PDF file.
PDFView online with eReader.
eReaderView this article in HTML Format.
HTML FormatCheck if you have access through your login credentials or your institution to get full access on this article.
Sign in