NVIDIA · Volta · 2017
NVIDIA V100
The first Tensor Core GPU. Retired from the frontier, still cheap and capable for small models.
32 GB
HBM2
0.9 TB/s
Mem bandwidth
125 TF
FP16 dense
300 W
TDP
$0.21 /hr
From · Vast.ai
No BF16 support complicates modern training recipes. As a cheap rental it remains serviceable for inference of models up to ~13B quantized.
Where to rent a V100
On-demand prices per GPU-hour; marketplace rates refresh daily, list prices reviewed 2026-09. Reserved and spot run 30-70% lower. Full breakdown with the used market and rent-vs-buy math: V100 pricing.
| Provider | $/GPU-hr | Notes |
|---|---|---|
| Vast.ai | $0.21 | Verified-listing rate, refreshed daily |
| RunPod | $0.29 | neocloud |
| OVHcloud | $0.88 | hyperscaler |
| AWS | $3.06 | p3.2xlarge, legacy pricing |
Specifications in context
Bars scaled against the best value in the whole catalog (B200 / MI325X era).
| Architecture | Volta (2017) |
|---|---|
| Memory | 32 GB HBM2, 0.9 TB/s |
| Interconnect | NVLink 2 · 300 GB/s |
| Form factor | SXM / PCIe |
| Partitioning | No MIG (time-slicing / vGPU only) |
What fits on one V100
Weights + ~2 GB runtime overhead against 29 GB usable VRAM. Longer contexts and bigger batches need more, check the VRAM calculator.
| Model | Params | Highest precision that fits |
|---|---|---|
| Llama 3.2 1B | 1.24B | FP16 |
| Llama 3.2 3B | 3.21B | FP16 |
| Llama 3.1 8B | 8.03B | FP16 |
| Qwen2.5 7B | 7.62B | FP16 |
| Qwen2.5 14B | 14.8B | 8-bit |
| Qwen2.5 32B | 32.8B | 4-bit |
| Qwen2.5 Coder 32B | 32.8B | 4-bit |
| QwQ 32B (reasoning) | 32.8B | 4-bit |
| Mistral 7B | 7.25B | FP16 |
| Gemma 2 9B | 9.24B | FP16 |
| Gemma 2 27B | 27.2B | 4-bit |
| Phi-4 14B | 14.7B | 8-bit |
| gpt-oss-20b | 20.9B | 8-bit |
Best for
- Legacy CUDA workloads
- Cheap experimentation
- Small-model inference