NVIDIA GH200 Superchip vs NVIDIA Tesla V100 — Paghahambing ng GPU (Jun 2026)
NVIDIA GH200 Superchip (96GB HBM3, 989 TFLOPS FP16, Hopper) vs NVIDIA Tesla V100 (16GB HBM2, 125 TFLOPS FP16, Volta). Cloud pricing: NVIDIA Tesla V100 from $0.13/hr. Ihambing ang specs, VRAM, performance, at presyo sa 1 cloud providers upang mahanap ang pinakamahusay na GPU para sa iyong AI workload.
|
NVIDIA GH200 Superchip
96GB HBM3 · Hopper
|
NVIDIA Tesla V100
16GB HBM2 · Volta
|
||
|---|---|---|---|
| Mga Espesipikasyon | |||
| Tagagawa | NVIDIA | NVIDIA | |
| Arkitektura | Hopper | Volta | |
| VRAM | 96 GB HBM3 | 16 GB HBM2 | |
| Bandwidth | 4,000 GB/s | 900 GB/s | |
| FP16 (Tensor) | 989.0 TFLOPS | 125.0 TFLOPS | |
| FP32 | 494.5 TFLOPS | 15.7 TFLOPS | |
| TDP | 700W | 300W | |
| Taon ng Paglabas | 2023 | 2017 | |
| Segmento | Data center | Data center | |
| Pinakamainam Para sa | Large-scale AI training HPC | Legacy training inference HPC | |
| Presyo sa Cloud | |||
| Pinakamurang On-Demand | — | $0.13/hr | |
| Pinakamurang Spot | — | — | |
| Mga Provider | 0 | 1 | |
| Presyo ng Provider (On-Demand) | |||
|
Hindi naaangkop | $0.13/hr | |
NVIDIA GH200 Superchip
NVIDIA Tesla V100