GPU × use-case guide · Code assistant
Is the H100 PCIe a good GPU for code assistant?
The H100 PCIe is a hopper NVIDIA GPU with 80 GB HBM3 (2000 GB/s), 756 TFLOPS BF16 / 1513 TFLOPS FP8, and a 350 W TDP. Code assistant workloads care most about low time-to-first-token and high single-stream tokens/sec, since developers wait on completions in real time. Here's how the H100 PCIe measures up.
What models fit on a single H100 PCIe?
Weights only, reserving ~25% of the 80 GB for KV cache, activations and fragmentation. ✓ = fits on one card.
| Model | BF16 | FP8 | INT4 |
|---|---|---|---|
| Llama 3.1 8B | ✓ | ✓ | ✓ |
| Qwen 2.5 14B | ✓ | ✓ | ✓ |
| Gemma 2 27B | ✓ | ✓ | ✓ |
| Mixtral 8x7B (MoE) | — | ✓ | ✓ |
| Llama 3.3 70B | — | — | ✓ |
| Qwen 2.5 72B | — | — | ✓ |
| Llama 3.1 405B | — | — | — |
Largest single-card fit: Gemma 2 27B at BF16, Mixtral 8x7B (MoE) at FP8, Qwen 2.5 72B at INT4. Bigger models need tensor-parallel across 8 cards.
H100 PCIe for code assistant, specifically
Code assistant is context-heavy, so the KV cache — not the weights — is what fills the 80 GB. On the H100 PCIe you'll trade context length against batch size: long prompts mean fewer concurrent requests. Because it's latency-sensitive, run it at low batch sizes on a fast framework (vLLM ≈ 85% MFU) rather than maximising batch. Size it precisely on the calculator.
H100 PCIe pricing across providers
| Provider | On-demand $/hr | Reserved $/hr |
|---|---|---|
| lambda | $2.29 | — |
| tensordock | $2.59 | — |
| vast_ai | $2.8 | — |
| runpod | $3.09 | — |
Verdict
With 80 GB, the H100 PCIe is a data-center-class card that comfortably handles code assistant for models up to Gemma 2 27B at full precision on a single card — a strong pick if your budget supports ~$2.29/hr.
See full H100 PCIespecs & pricing, size your model on the calculator, or compare every GPU on the GPU list.