Benchmarks 2024 03 02 TVM LLVM Os - tum-ei-eda/muriscv-nn GitHub Wiki
Setup
Simulator
- Spike (
riscv-isa-sim) (ISS, CPI=1)
Toolchains
- LLVM/Clang:
- TODO: Version
- Linker: lld (TODO)
- RISC-V GCC for Headers, libc,...
Models
-
MLPerfTiny Benchmark
-
TODO: others!
Package Versions
-
MLonMCU : main
-
TVM : Nightly Pre-Build
-
Spike : 0bc176b3fca43560b9e8586cdbc41cfde073e17a
-
Spike PK : 7e9b671c0415dfd7b562ac934feb9380075d4aa2
Miscellaneous
- Used
-Osflag for compilation. - Benchmarks generated using MLonMCU deployment tool with minimal efforts.
- Memory metrics are reported in Bytes
Results (Framework: tvm, Backend: tvmaot, Toolchain: llvm, Flags: -Os)
Audio Wake Words (aww)
| Cycles (Speedup) | Total ROM (rel.) | Total RAM (rel.) | VLEN | Layout | Kernels | Mode | Arch | Unroll | Auto-Vectorization |
|---|---|---|---|---|---|---|---|---|---|
33508604 ( 0.5x ) |
109110 ( 1.202 ) |
59508 ( 3.097 ) |
0 | NCHW | TVM | Fallback | RV32GC | 0 | - |
27511073 ( 0.6x ) |
102506 ( 1.13 ) |
59508 ( 3.097 ) |
0 | NHWC | TVM | Fallback | RV32GC | 0 | - |
13706216 ( 1.1x ) |
102504 ( 1.13 ) |
51336 ( 2.672 ) |
0 | NCHW | TVM | Autotuned | RV32GC | 0 | - |
27505623 ( 0.6x ) |
102618 ( 1.131 ) |
59508 ( 3.097 ) |
0 | NHWC | TVM | Autotuned | RV32GC | 0 | - |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
128 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
256 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
512 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
1024 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
2048 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
3384606 ( 4.6x ) |
105404 ( 1.162 ) |
59508 ( 3.097 ) |
4096 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
9566669 ( 1.6x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
128 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
7918669 ( 2.0x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
256 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
7094669 ( 2.2x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
512 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
6682669 ( 2.3x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
1024 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
6682669 ( 2.3x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
2048 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
6682669 ( 2.3x ) |
103606 ( 1.142 ) |
59508 ( 3.097 ) |
4096 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
5607715 ( 2.8x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
128 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
4512654 ( 3.5x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
256 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
4089948 ( 3.8x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
512 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
3881984 ( 4.0x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
1024 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
3881984 ( 4.0x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
2048 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
3888762 ( 4.0x ) |
106776 ( 1.177 ) |
51336 ( 2.672 ) |
4096 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
9565134 ( 1.6x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
128 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
7916473 ( 2.0x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
256 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
7092167 ( 2.2x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
512 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
6683403 ( 2.3x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
1024 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
6683403 ( 2.3x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
2048 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
6690181 ( 2.3x ) |
104544 ( 1.152 ) |
59508 ( 3.097 ) |
4096 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
15615253 ( Base ) |
90744 ( Base ) |
19212 ( Base ) |
0 | NHWC | muRISCV-NN | Scalar | RV32GC | 0 | - |
6838472 ( 2.3x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5996140 ( 2.6x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5575084 ( 2.8x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5367945 ( 2.9x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5367945 ( 2.9x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5374723 ( 2.9x ) |
93956 ( 1.035 ) |
19212 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
5187254 ( 3.0x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
128 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
4047492 ( 3.9x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
256 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
3486214 ( 4.5x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
512 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
3443726 ( 4.5x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
1024 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
3447115 ( 4.5x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
2048 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
3450504 ( 4.5x ) |
91052 ( 1.003 ) |
23676 ( 1.232 ) |
4096 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
Image Classification (resnet)
| Cycles (Speedup) | Total ROM (rel.) | Total RAM (rel.) | VLEN | Layout | Kernels | Mode | Arch | Unroll | Auto-Vectorization |
|---|---|---|---|---|---|---|---|---|---|
144789098 ( 0.4x ) |
218264 ( 1.579 ) |
108420 ( 1.953 ) |
0 | NCHW | TVM | Fallback | RV32GC | 0 | - |
112394400 ( 0.5x ) |
209174 ( 1.514 ) |
108420 ( 1.953 ) |
0 | NHWC | TVM | Fallback | RV32GC | 0 | - |
53701817 ( 1.1x ) |
212582 ( 1.538 ) |
92236 ( 1.661 ) |
0 | NCHW | TVM | Autotuned | RV32GC | 0 | - |
112389724 ( 0.5x ) |
209256 ( 1.514 ) |
108420 ( 1.953 ) |
0 | NHWC | TVM | Autotuned | RV32GC | 0 | - |
12825993 ( 4.6x ) |
213686 ( 1.546 ) |
108420 ( 1.953 ) |
128 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
12825992 ( 4.6x ) |
213684 ( 1.546 ) |
108420 ( 1.953 ) |
256 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
12825992 ( 4.6x ) |
213684 ( 1.546 ) |
108420 ( 1.953 ) |
512 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
12825991 ( 4.6x ) |
213678 ( 1.546 ) |
108420 ( 1.953 ) |
1024 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
12825990 ( 4.6x ) |
213676 ( 1.546 ) |
108420 ( 1.953 ) |
2048 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
12825990 ( 4.6x ) |
213676 ( 1.546 ) |
108420 ( 1.953 ) |
4096 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
36071697 ( 1.6x ) |
210202 ( 1.521 ) |
108420 ( 1.953 ) |
128 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
27020559 ( 2.2x ) |
210194 ( 1.521 ) |
108420 ( 1.953 ) |
256 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
24755983 ( 2.4x ) |
210194 ( 1.521 ) |
108420 ( 1.953 ) |
512 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
24311825 ( 2.4x ) |
210202 ( 1.521 ) |
108420 ( 1.953 ) |
1024 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
24311823 ( 2.4x ) |
210194 ( 1.521 ) |
108420 ( 1.953 ) |
2048 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
24311823 ( 2.4x ) |
210194 ( 1.521 ) |
108420 ( 1.953 ) |
4096 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
19686220 ( 3.0x ) |
226232 ( 1.637 ) |
92236 ( 1.661 ) |
128 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
15460448 ( 3.8x ) |
226232 ( 1.637 ) |
92236 ( 1.661 ) |
256 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
14238174 ( 4.1x ) |
226232 ( 1.637 ) |
92236 ( 1.661 ) |
512 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
13965590 ( 4.2x ) |
226230 ( 1.637 ) |
92236 ( 1.661 ) |
1024 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
13982536 ( 4.2x ) |
226232 ( 1.637 ) |
92236 ( 1.661 ) |
2048 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
14023205 ( 4.2x ) |
226234 ( 1.637 ) |
92236 ( 1.661 ) |
4096 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
36069956 ( 1.6x ) |
210938 ( 1.526 ) |
108420 ( 1.953 ) |
128 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
27018265 ( 2.2x ) |
210930 ( 1.526 ) |
108420 ( 1.953 ) |
256 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
24753437 ( 2.4x ) |
210930 ( 1.526 ) |
108420 ( 1.953 ) |
512 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
24309151 ( 2.4x ) |
210930 ( 1.526 ) |
108420 ( 1.953 ) |
1024 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
24312542 ( 2.4x ) |
210938 ( 1.526 ) |
108420 ( 1.953 ) |
2048 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
24315931 ( 2.4x ) |
210938 ( 1.526 ) |
108420 ( 1.953 ) |
4096 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
58407308 ( Base ) |
138192 ( Base ) |
55516 ( Base ) |
0 | NHWC | muRISCV-NN | Scalar | RV32GC | 0 | - |
28255208 ( 2.1x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
19603964 ( 3.0x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
15548788 ( 3.8x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
13704813 ( 4.3x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
12803693 ( 4.6x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
12134631 ( 4.8x ) |
141914 ( 1.027 ) |
55516 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
15346885 ( 3.8x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
9761897 ( 6.0x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
7215163 ( 8.1x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
5972587 ( 9.8x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
5047488 ( 11.6x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
4801673 ( 12.2x ) |
139140 ( 1.007 ) |
55516 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
Anomaly Detection (toycar)
| Cycles (Speedup) | Total ROM (rel.) | Total RAM (rel.) | VLEN | Layout | Kernels | Mode | Arch | Unroll | Auto-Vectorization |
|---|---|---|---|---|---|---|---|---|---|
3404882 ( 0.5x ) |
581370 ( 1.84 ) |
5572 ( 1.168 ) |
0 | NCHW | TVM | Fallback | RV32GC | 0 | - |
3404880 ( 0.5x ) |
581362 ( 1.84 ) |
5572 ( 1.168 ) |
0 | NHWC | TVM | Fallback | RV32GC | 0 | - |
2245737 ( 0.8x ) |
609080 ( 1.928 ) |
6884 ( 1.443 ) |
0 | NCHW | TVM | Autotuned | RV32GC | 0 | - |
2245737 ( 0.8x ) |
609080 ( 1.928 ) |
6884 ( 1.443 ) |
0 | NHWC | TVM | Autotuned | RV32GC | 0 | - |
984695 ( 1.8x ) |
581106 ( 1.839 ) |
5572 ( 1.168 ) |
128 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
256 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
512 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984695 ( 1.8x ) |
581106 ( 1.839 ) |
5572 ( 1.168 ) |
1024 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
2048 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984695 ( 1.8x ) |
581106 ( 1.839 ) |
5572 ( 1.168 ) |
4096 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
128 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
256 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
512 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
1024 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984693 ( 1.8x ) |
581098 ( 1.839 ) |
5572 ( 1.168 ) |
2048 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
984695 ( 1.8x ) |
581106 ( 1.839 ) |
5572 ( 1.168 ) |
4096 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
1280619 ( 1.3x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
128 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1202747 ( 1.4x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
256 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1164011 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
512 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1148032 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
1024 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1145126 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
2048 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1151904 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
4096 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1280619 ( 1.3x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
128 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1202747 ( 1.4x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
256 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1164011 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
512 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1148032 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
1024 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1145126 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
2048 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1151904 ( 1.5x ) |
600432 ( 1.9 ) |
6884 ( 1.443 ) |
4096 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
1724219 ( Base ) |
315976 ( Base ) |
4772 ( Base ) |
0 | NHWC | muRISCV-NN | Scalar | RV32GC | 0 | - |
620488 ( 2.8x ) |
316614 ( 1.002 ) |
4772 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
506567 ( 3.4x ) |
316612 ( 1.002 ) |
4772 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
449606 ( 3.8x ) |
316610 ( 1.002 ) |
4772 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
421128 ( 4.1x ) |
316616 ( 1.002 ) |
4772 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
406888 ( 4.2x ) |
316614 ( 1.002 ) |
4772 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
405102 ( 4.3x ) |
316610 ( 1.002 ) |
4772 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
618525 ( 2.8x ) |
316516 ( 1.002 ) |
4772 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
501848 ( 3.4x ) |
316526 ( 1.002 ) |
4772 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
443506 ( 3.9x ) |
316518 ( 1.002 ) |
4772 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
414659 ( 4.2x ) |
316516 ( 1.002 ) |
4772 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
411029 ( 4.2x ) |
316518 ( 1.002 ) |
4772 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
409172 ( 4.2x ) |
316516 ( 1.002 ) |
4772 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
Visual Wake Words (vww)
| Cycles (Speedup) | Total ROM (rel.) | Total RAM (rel.) | VLEN | Layout | Kernels | Mode | Arch | Unroll | Auto-Vectorization |
|---|---|---|---|---|---|---|---|---|---|
96665971 ( 0.5x ) |
545174 ( 1.684 ) |
181032 ( 2.113 ) |
0 | NCHW | TVM | Fallback | RV32GC | 0 | - |
79940189 ( 0.6x ) |
521128 ( 1.61 ) |
181032 ( 2.113 ) |
0 | NHWC | TVM | Fallback | RV32GC | 0 | - |
42404607 ( 1.1x ) |
525218 ( 1.622 ) |
181032 ( 2.113 ) |
0 | NCHW | TVM | Autotuned | RV32GC | 0 | - |
79940191 ( 0.6x ) |
521128 ( 1.61 ) |
181032 ( 2.113 ) |
0 | NHWC | TVM | Autotuned | RV32GC | 0 | - |
11010120 ( 4.2x ) |
532510 ( 1.645 ) |
181032 ( 2.113 ) |
128 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
11010119 ( 4.2x ) |
532508 ( 1.645 ) |
181032 ( 2.113 ) |
256 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
11010118 ( 4.2x ) |
532506 ( 1.645 ) |
181032 ( 2.113 ) |
512 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
11010120 ( 4.2x ) |
532510 ( 1.645 ) |
181032 ( 2.113 ) |
1024 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
11010121 ( 4.2x ) |
532512 ( 1.645 ) |
181032 ( 2.113 ) |
2048 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
11010121 ( 4.2x ) |
532512 ( 1.645 ) |
181032 ( 2.113 ) |
4096 | NCHW | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
30451801 ( 1.5x ) |
523638 ( 1.617 ) |
181032 ( 2.113 ) |
128 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
25820476 ( 1.8x ) |
523640 ( 1.617 ) |
181032 ( 2.113 ) |
256 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
23623465 ( 2.0x ) |
523638 ( 1.617 ) |
181032 ( 2.113 ) |
512 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
22700929 ( 2.1x ) |
523636 ( 1.617 ) |
181032 ( 2.113 ) |
1024 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
22335710 ( 2.1x ) |
523638 ( 1.617 ) |
181032 ( 2.113 ) |
2048 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
22307810 ( 2.1x ) |
523638 ( 1.617 ) |
181032 ( 2.113 ) |
4096 | NHWC | TVM | Fallback | RV32GCV | 0 | Loop+SLP |
24965509 ( 1.9x ) |
550714 ( 1.701 ) |
181032 ( 2.113 ) |
128 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
21454905 ( 2.2x ) |
550730 ( 1.701 ) |
181032 ( 2.113 ) |
256 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
19785359 ( 2.4x ) |
550732 ( 1.701 ) |
181032 ( 2.113 ) |
512 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
19204884 ( 2.4x ) |
550730 ( 1.701 ) |
181032 ( 2.113 ) |
1024 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
19019631 ( 2.5x ) |
550738 ( 1.701 ) |
181032 ( 2.113 ) |
2048 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
19039948 ( 2.5x ) |
550734 ( 1.701 ) |
181032 ( 2.113 ) |
4096 | NCHW | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
30450316 ( 1.5x ) |
523746 ( 1.618 ) |
181032 ( 2.113 ) |
128 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
25818717 ( 1.8x ) |
523748 ( 1.618 ) |
181032 ( 2.113 ) |
256 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
23621575 ( 2.0x ) |
523750 ( 1.618 ) |
181032 ( 2.113 ) |
512 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
22698971 ( 2.1x ) |
523750 ( 1.618 ) |
181032 ( 2.113 ) |
1024 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
22333715 ( 2.1x ) |
523748 ( 1.618 ) |
181032 ( 2.113 ) |
2048 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
22305797 ( 2.1x ) |
523748 ( 1.618 ) |
181032 ( 2.113 ) |
4096 | NHWC | TVM | Autotuned | RV32GCV | 0 | Loop+SLP |
46765906 ( Base ) |
323760 ( Base ) |
85664 ( Base ) |
0 | NHWC | muRISCV-NN | Scalar | RV32GC | 0 | - |
19684203 ( 2.4x ) |
327944 ( 1.013 ) |
85664 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
16780939 ( 2.8x ) |
327942 ( 1.013 ) |
85664 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
15389323 ( 3.0x ) |
327944 ( 1.013 ) |
85664 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
14888472 ( 3.1x ) |
327942 ( 1.013 ) |
85664 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
14681952 ( 3.2x ) |
327944 ( 1.013 ) |
85664 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
14671295 ( 3.2x ) |
327948 ( 1.013 ) |
85664 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Scalar | RV32GCV | 0 | Loop+SLP |
13749962 ( 3.4x ) |
325152 ( 1.004 ) |
85664 ( 1.0 ) |
128 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
10417322 ( 4.5x ) |
325150 ( 1.004 ) |
85664 ( 1.0 ) |
256 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
9120938 ( 5.1x ) |
325150 ( 1.004 ) |
85664 ( 1.0 ) |
512 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
8615422 ( 5.4x ) |
325152 ( 1.004 ) |
85664 ( 1.0 ) |
1024 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
8570421 ( 5.5x ) |
325150 ( 1.004 ) |
85664 ( 1.0 ) |
2048 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
8573810 ( 5.5x ) |
325152 ( 1.004 ) |
85664 ( 1.0 ) |
4096 | NHWC | muRISCV-NN | Vector | RV32GCV | 0 | - |
Original data
Click here to download the raw files for this benchmark.