GPU Workload Scorer
Not all GPUs are equal for every task. We score every GPU in our catalog across five real AI workloads using weighted scoring on core specs — VRAM, FP16 TFLOPS, memory bandwidth, and TDP.
Training weights VRAM and compute highest. Inference prioritizes memory bandwidth and throughput. Edge deployment optimizes for low power. Select a workload to see the rankings.
Choose a workload profile
LLM Training ranking
| Rank | GPU | Score |
|---|---|---|
| #1 | NVIDIA A100 Ampere | Excellent 97% |
| #2 | AMD Instinct MI250X CDNA 2 | Excellent 95% |
| #3 | NVIDIA H100 Hopper | Excellent 95% |
| #4 | NVIDIA H200 Hopper | Excellent 95% |
| #5 | NVIDIA H200 NVL Hopper | Excellent 95% |
| #6 | NVIDIA B200 Blackwell | Excellent 95% |
| #7 | NVIDIA B300 Blackwell Ultra | Excellent 95% |
| #8 | NVIDIA GB200 Blackwell | Excellent 95% |
| #9 | NVIDIA GH200 Hopper | Excellent 95% |
| #10 | AMD Instinct MI300X CDNA 3 | Excellent 95% |
| #11 | NVIDIA RTX 6000 Ada Ada Lovelace | Good 72% |
| #12 | NVIDIA L40S Ada Lovelace | Good 71% |
| #13 | NVIDIA RTX A6000 Ampere | Good 63% |
| #14 | NVIDIA A40 Ampere | Good 61% |
| #15 | NVIDIA A16 Ampere | Fair 53% |