MapIdTransform Workload Latency Comparison
0.039PyTorch Latency (ms)PyTorch
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 20002025.12 | 0.039 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 40002025.12 | 0.046 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 60002025.12 | 0.049 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 80002025.12 | 0.053 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=5000 × 100002025.12 | 0.06 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=100 × 100002025.12 | 0.061 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=500 × 100002025.12 | 0.063 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=1000 × 100002025.12 | 0.063 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 100002025.12 | 0.063 | — | — | |
| PyTorchAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 500002025.12 | 0.14 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 20002025.12 | 0.399 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 40002025.12 | 0.72 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 60002025.12 | 1.046 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 80002025.12 | 1.367 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=100 × 100002025.12 | 1.623 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=500 × 100002025.12 | 1.636 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=1000 × 100002025.12 | 1.641 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=5000 × 100002025.12 | 1.667 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 100002025.12 | 1.688 | — | — | |
| PyTorchAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 500002025.12 | 8.09 | — | — | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=100 × 100002025.12 | — | 0.466 | 3.48 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=500 × 100002025.12 | — | 0.472 | 3.47 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=1000 × 100002025.12 | — | 0.48 | 3.42 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=5000 × 100002025.12 | — | 0.508 | 3.28 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 20002025.12 | — | 0.514 | 0.78 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 40002025.12 | — | 0.521 | 1.38 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 60002025.12 | — | 0.523 | 2 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 80002025.12 | — | 0.52 | 2.63 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 100002025.12 | — | 0.523 | 3.23 | |
| KernelEvolveAI Hardware=MTIA v2i, Tensor Shape (UniqueIDs × Batch)=10000 × 500002025.12 | — | 1.989 | 4.07 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=100 × 100002025.12 | — | 0.058 | 1.05 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=500 × 100002025.12 | — | 0.055 | 1.15 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=1000 × 100002025.12 | — | 0.05 | 1.26 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=5000 × 100002025.12 | — | 0.048 | 1.25 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 20002025.12 | — | 0.035 | 1.11 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 40002025.12 | — | 0.036 | 1.28 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 60002025.12 | — | 0.037 | 1.32 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 80002025.12 | — | 0.039 | 1.36 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 100002025.12 | — | 0.048 | 1.31 | |
| KernelEvolveAI Hardware=MTIA v3, Tensor Shape (UniqueIDs × Batch)=10000 × 500002025.12 | — | 0.174 | 0.8 |