Matrix Multiplication Performance on RTX 4090 (Sizes 2048-8192)
0.1002Latency (ms)RT-Lynx
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RT-LynxMatrix Shape (M=N)=2048, Matrix Shape (K)=30722026.05 | 0.1002 | 1.61 | 10.97 | |
| GEMMMatrix Shape (M=N)=2048, Matrix Shape (K)=30722026.05 | 0.1618 | — | — | |
| cuSparseLtMatrix Shape (M=N)=2048, Matrix Shape (K)=30722026.05 | 0.2408 | 0.67 | 64.04 | |
| RT-LynxMatrix Shape (M=N)=4096, Matrix Shape (K)=30722026.05 | 0.354 | 1.78 | 5.79 | |
| RT-LynxMatrix Shape (M=N)=2048, Matrix Shape (K)=122882026.05 | 0.4003 | 1.67 | 16.64 | |
| Pytorch-SpMMMatrix Shape (M=N)=2048, Matrix Shape (K)=30722026.05 | 0.5226 | 0.31 | 38.18 | |
| cuSparseLtMatrix Shape (M=N)=4096, Matrix Shape (K)=30722026.05 | 0.5956 | 1.06 | 47.61 | |
| Pytorch-SpMMMatrix Shape (M=N)=4096, Matrix Shape (K)=30722026.05 | 0.6053 | 1.04 | 44.31 | |
| GEMMMatrix Shape (M=N)=4096, Matrix Shape (K)=30722026.05 | 0.6301 | — | — | |
| GEMMMatrix Shape (M=N)=2048, Matrix Shape (K)=122882026.05 | 0.667 | — | — | |
| Pytorch-SpMMMatrix Shape (M=N)=2048, Matrix Shape (K)=122882026.05 | 0.885 | 0.75 | 63.24 | |
| cuSparseLtMatrix Shape (M=N)=2048, Matrix Shape (K)=122882026.05 | 0.9122 | 0.73 | 66.72 | |
| RT-LynxMatrix Shape (M=N)=8192, Matrix Shape (K)=30722026.05 | 1.4466 | 1.79 | 4.38 | |
| RT-LynxMatrix Shape (M=N)=4096, Matrix Shape (K)=122882026.05 | 1.5198 | 1.73 | 13.09 | |
| Pytorch-SpMMMatrix Shape (M=N)=8192, Matrix Shape (K)=30722026.05 | 1.8729 | 1.38 | 29.43 | |
| cuSparseLtMatrix Shape (M=N)=8192, Matrix Shape (K)=30722026.05 | 1.9869 | 1.3 | 30.51 | |
| Pytorch-SpMMMatrix Shape (M=N)=4096, Matrix Shape (K)=122882026.05 | 2.3781 | 1.11 | 47.85 | |
| cuSparseLtMatrix Shape (M=N)=4096, Matrix Shape (K)=122882026.05 | 2.5827 | 1.02 | 52.84 | |
| GEMMMatrix Shape (M=N)=8192, Matrix Shape (K)=30722026.05 | 2.585 | — | — | |
| GEMMMatrix Shape (M=N)=4096, Matrix Shape (K)=122882026.05 | 2.6281 | — | — | |
| RT-LynxMatrix Shape (M=N)=8192, Matrix Shape (K)=122882026.05 | 6.26 | 1.69 | 6.41 | |
| Pytorch-SpMMMatrix Shape (M=N)=8192, Matrix Shape (K)=122882026.05 | 7.4751 | 1.41 | 29.8 | |
| cuSparseLtMatrix Shape (M=N)=8192, Matrix Shape (K)=122882026.05 | 7.6654 | 1.38 | 35.18 | |
| GEMMMatrix Shape (M=N)=8192, Matrix Shape (K)=122882026.05 | 10.5636 | — | — |