Traffic Forecasting Efficiency on NVIDIA V100 GPU Configuration
3.8Parameters (M)DCRNN
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DCRNNBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 3.8 | 2.1 | 28.4 | |
| STGCNBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 4.2 | 1.8 | 24.6 | |
| Graph WaveNetBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 5.1 | 2.4 | 31.2 | |
| CIWI-CKTBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 5.3 | 2.6 | 22.4 | |
| FlashSTBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 6.2 | 3.5 | 42.6 | |
| ST-GFSLBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 8.4 | 4.2 | 51.8 | |
| TransGTRBatch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 12.6 | 6.8 | 78.4 | |
| Transformer (4-layer)Batch Size (B)=32, Time steps (T)=96, Input Dimension (D)=16, Hardware=NVIDIA V100 32GB GPU2026.06 | 32.8 | 15.7 | 142.3 |