NVIDIA GH200 96GBBudget pickB75.9 tok/s decode
Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
40.1 tok/s
TTFT
4827 ms
Safe context
20K
Memory
62.8 GB / 80.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Tight fit | 40.1 tok/s | 7020 ms | 35K |
| Chat | C | Runs well | 40.1 tok/s | 2633 ms | 11K |
| Coding | C | Runs well | 40.1 tok/s | 4827 ms | 20K |
| RAG | C | Tight fit | 40.1 tok/s | 8776 ms | 35K |
| Reasoning | C | Runs well | 40.1 tok/s | 5704 ms | 20K |
How Llama 3.1 70B (70B params) fits at each quantization level on NVIDIA A100 80GB (80.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 27.3 GB | Low | D37 |
Q3_K_S | 3 | 34.3 GB | Low | D39 |
NVFP4 | 4 | 39.2 GB | Medium | D40 |
Q4_K_M | 4 | 42.7 GB | Medium | C41 |
Q5_K_M | 5 | 50.4 GB | High | C43 |
Q6_KBest for your GPU | 6 | 57.4 GB | High | C44 |
Q8_0 | 8 | 74.9 GB | Very High | C44 |
F16 | 16 | 143.5 GB | Maximum | F0 |
ollama run llama-3.1-70bhuggingface-cli download llama-3.1-70bUpgrade options
NVIDIA GH200 96GBBudget pick
NVIDIA H20 96GBBiggest leap