DeepSeek on NVIDIA Ampere
Can I run DeepSeek-R1 671B-A37B on an A100 80GB?
Not at Q4_K_M — it needs 380.4 GB against 74.4 GB available. You would need 6 of these cards.
Won't fit
380.4 GB of 74.4 GB · 511%0396 GB
Weights
379.0 GB
KV cache
0.5 GB
Runtime overhead
0.9 GB
Over the limit
306.0 GB
Short by 306.0 GB. You can run it with 11 of 61 layers on the A100 80GB and the rest in system RAM, at roughly 2.26 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.
Generation2.26tok/s
Prompt processing1771tok/s
Max context0tokens
KV per 1K tokens0GB
Every quantisation of DeepSeek-R1 671B-A37B on a A100 80GB
Highlighted row is the highest quality that still fits at 8K context.
| Quantisation | bpw | Weights | Total | Verdict | Max ctx | tok/s |
|---|---|---|---|---|---|---|
| FP16 / BF16 | 16.00 | 1249.8 GB | 1251.2 GB | Won't fit | — | 0.60 |
| INT8 / W8A8 | 8.50 | 663.6 GB | 665.0 GB | Won't fit | — | 1.19 |
| Q8_0 (GGUF) | 8.50 | 663.6 GB | 665.0 GB | Won't fit | — | 1.19 |
| FP8 (E4M3) | 8.00 | 624.9 GB | 626.3 GB | Won't fit | — | 1.28 |
| Q6_K | 6.56 | 512.4 GB | 513.8 GB | Won't fit | — | 1.59 |
| Q5_K_M | 5.67 | 442.9 GB | 444.3 GB | Won't fit | — | 1.90 |
| Q5_K_S | 5.52 | 431.2 GB | 432.6 GB | Won't fit | — | 1.95 |
| Q4_K_M | 4.85 | 379.0 GB | 380.4 GB | Won't fit | — | 2.26 |
| Q4_K_S | 4.58 | 358.0 GB | 359.4 GB | Won't fit | — | 2.43 |
| Q4_0 | 4.55 | 355.6 GB | 357.0 GB | Won't fit | — | 2.45 |
| AWQ 4-bit | 4.25 | 334.5 GB | 335.9 GB | Won't fit | — | 2.65 |
| GPTQ 4-bit | 4.25 | 334.5 GB | 335.9 GB | Won't fit | — | 2.65 |
| MXFP4 | 4.25 | 334.5 GB | 335.9 GB | Won't fit | — | 2.65 |
| IQ4_XS | 4.25 | 332.3 GB | 333.7 GB | Won't fit | — | 2.67 |
| Q3_K_M | 3.91 | 305.8 GB | 307.2 GB | Won't fit | — | 2.95 |
| IQ3_M | 3.70 | 289.4 GB | 290.8 GB | Won't fit | — | 3.17 |
| IQ3_XXS | 3.06 | 239.6 GB | 241.0 GB | Won't fit | — | 4.06 |
| Q2_K | 2.63 | 206.1 GB | 207.5 GB | Won't fit | — | 5.01 |
| IQ2_XXS | 2.06 | 161.7 GB | 163.1 GB | Won't fit | — | 7.30 |
| IQ1_M | 1.75 | 137.5 GB | 138.9 GB | Won't fit | — | 9.77 |