Mistral on NVIDIA Ampere
Can I run Mixtral 8x7B on an RTX 3090?
Not at Q4_K_M — it needs 28.2 GB against 21.8 GB available. Drop to IQ3_XXS and it fits, at about 92.3 tokens per second.
Won't fit
28.2 GB of 21.8 GB · 130%029 GB
Weights
26.4 GB
KV cache
1.0 GB
Runtime overhead
0.8 GB
Over the limit
6.4 GB
Short by 6.4 GB. You can run it with 24 of 32 layers on the RTX 3090 and the rest in system RAM, at roughly 16.4 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.
Generation16.4tok/s
Prompt processing1156tok/s
Max context0tokens
KV per 1K tokens0GB
Every quantisation of Mixtral 8x7B on a RTX 3090
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 | 87.0 GB | 88.8 GB | Won't fit | — | 2.04 |
| INT8 / W8A8 | 8.50 | 46.2 GB | 48.0 GB | Won't fit | — | 4.80 |
| Q8_0 (GGUF) | 8.50 | 46.2 GB | 48.0 GB | Won't fit | — | 4.80 |
| FP8 (E4M3) | 8.00 | 43.5 GB | 45.3 GB | Won't fit | — | 5.33 |
| Q6_K | 6.56 | 35.7 GB | 37.5 GB | Won't fit | — | 7.52 |
| Q5_K_M | 5.67 | 30.8 GB | 32.6 GB | Won't fit | — | 10.4 |
| Q5_K_S | 5.52 | 30.0 GB | 31.8 GB | Won't fit | — | 11.4 |
| Q4_K_M | 4.85 | 26.4 GB | 28.2 GB | Won't fit | — | 16.4 |
| Q4_K_S | 4.58 | 24.9 GB | 26.7 GB | Won't fit | — | 19.0 |
| Q4_0 | 4.55 | 24.8 GB | 26.6 GB | Won't fit | — | 19.1 |
| AWQ 4-bit | 4.25 | 23.5 GB | 25.3 GB | Won't fit | — | 25.2 |
| GPTQ 4-bit | 4.25 | 23.5 GB | 25.3 GB | Won't fit | — | 25.2 |
| MXFP4 | 4.25 | 23.5 GB | 25.3 GB | Won't fit | — | 25.2 |
| IQ4_XS | 4.25 | 23.1 GB | 25.0 GB | Won't fit | — | 25.5 |
| Q3_K_M | 3.91 | 21.3 GB | 23.1 GB | Won't fit | — | 36.8 |
| IQ3_M | 3.70 | 20.2 GB | 22.0 GB | Won't fit | 6K | 58.5 |
| IQ3_XXS | 3.06 | 16.7 GB | 18.5 GB | Runs comfortably | 32K | 92.3 |
| Q2_K | 2.63 | 14.4 GB | 16.2 GB | Runs comfortably | 32K | 105 |
| IQ2_XXS | 2.06 | 11.3 GB | 13.1 GB | Runs comfortably | 32K | 127 |
| IQ1_M | 1.75 | 9.6 GB | 11.4 GB | Runs comfortably | 32K | 144 |