Jacobian computation on 2-layer MLP
0.29Median Runtime (ms)AlphaGrad
Evaluation Results
| Method | Links | |
|---|---|---|
| AlphaGradAD Mode=AlphaGrad, Device=NVIDIA RTX 4090 (GPU), Batch Size=512, Compilation=JIT2024.06 | 0.29 | |
| ReverseAD Mode=Reverse, Device=NVIDIA RTX 4090 (GPU), Batch Size=512, Compilation=JIT2024.06 | 0.3 | |
| MarkowitzAD Mode=Markowitz, Device=NVIDIA RTX 4090 (GPU), Batch Size=512, Compilation=JIT2024.06 | 1.01 | |
| ForwardAD Mode=Forward, Device=NVIDIA RTX 4090 (GPU), Batch Size=512, Compilation=JIT2024.06 | 11.04 | |
| AlphaGradAD Mode=AlphaGrad, Device=AMD EPYC 9684X (CPU), Batch Size=512, Compilation=XLA2024.06 | 28.73 | |
| ReverseAD Mode=Reverse, Device=AMD EPYC 9684X (CPU), Batch Size=512, Compilation=XLA2024.06 | 29.67 | |
| MarkowitzAD Mode=Markowitz, Device=AMD EPYC 9684X (CPU), Batch Size=512, Compilation=XLA2024.06 | 317.65 | |
| ForwardAD Mode=Forward, Device=AMD EPYC 9684X (CPU), Batch Size=512, Compilation=XLA2024.06 | 760.63 |