NVIDIA A2 vs NVIDIA GB200 Superchip — Comparaison GPU (Apr 2026)
NVIDIA A2 (16GB GDDR6, 18 TFLOPS FP16, Ampere) vs NVIDIA GB200 Superchip (384GB HBM3e, 4,500 TFLOPS FP16, Blackwell). Cloud pricing: NVIDIA A2 from $0.22/hr. Compare specs, VRAM, performance, and pricing across 1 cloud providers to find the best GPU for your AI workload.
|
NVIDIA A2
16GB GDDR6 · Ampere
|
NVIDIA GB200 Superchip
384GB HBM3e · Blackwell
|
||
|---|---|---|---|
| Spécifications | |||
| Fabricant | NVIDIA | NVIDIA | |
| Architecture | Ampere | Blackwell | |
| VRAM | 16 GB GDDR6 | 384 GB HBM3e | |
| Bande passante | 200 GB/s | 16,000 GB/s | |
| FP16 (Tensor) | 18.0 TFLOPS | 4500.0 TFLOPS | |
| FP32 | 4.5 TFLOPS | 150.0 TFLOPS | |
| TDP | 60W | 2700W | |
| Année de sortie | 2021 | 2024 | |
| Segment | Data center | Data center | |
| Idéal pour | Edge inference entry-level AI | Largest-scale AI training multi-trillion parameter models | |
| Tarification Cloud | |||
| Le moins cher à la demande | $0.22/hr | — | |
| Le moins cher en Spot | — | — | |
| Fournisseurs | 1 | 0 | |
| Tarification Fournisseur (À la demande) | |||
|
$0.22/hr | N/A | |
NVIDIA A2
NVIDIA GB200 Superchip