Llama on NVIDIA Ada
Can I run Llama 4 Scout 109B-A17B on an RTX 4070 Ti Super?
Not at Q4_K_M — it needs 64.0 GB against 14.2 GB available. You would need 5 of these cards.
Won't fit
64.0 GB of 14.2 GB · 450%067 GB
Weights
61.7 GB
KV cache
1.5 GB
Runtime overhead
0.8 GB
Over the limit
49.8 GB
Short by 49.8 GB. You can run it with 9 of 48 layers on the RTX 4070 Ti Super and the rest in system RAM, at roughly 4.42 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.
Generation4.42tok/s
Prompt processing1087tok/s
Max context0tokens
KV per 1K tokens0GB
Every quantisation of Llama 4 Scout 109B-A17B on a RTX 4070 Ti Super
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 | 203.0 GB | 205.4 GB | Won't fit | — | 1.23 |
| INT8 / W8A8 | 8.50 | 107.4 GB | 109.7 GB | Won't fit | — | 2.41 |
| Q8_0 (GGUF) | 8.50 | 107.4 GB | 109.7 GB | Won't fit | — | 2.41 |
| FP8 (E4M3) | 8.00 | 101.5 GB | 103.9 GB | Won't fit | — | 2.54 |
| Q6_K | 6.56 | 83.2 GB | 85.6 GB | Won't fit | — | 3.13 |
| Q5_K_M | 5.67 | 71.9 GB | 74.3 GB | Won't fit | — | 3.67 |
| Q5_K_S | 5.52 | 70.0 GB | 72.4 GB | Won't fit | — | 3.84 |
| Q4_K_M | 4.85 | 61.7 GB | 64.0 GB | Won't fit | — | 4.42 |
| Q4_K_S | 4.58 | 58.3 GB | 60.7 GB | Won't fit | — | 4.65 |
| Q4_0 | 4.55 | 58.0 GB | 60.3 GB | Won't fit | — | 4.68 |
| AWQ 4-bit | 4.25 | 56.8 GB | 59.1 GB | Won't fit | — | 4.88 |
| GPTQ 4-bit | 4.25 | 56.8 GB | 59.1 GB | Won't fit | — | 4.88 |
| MXFP4 | 4.25 | 56.8 GB | 59.1 GB | Won't fit | — | 4.88 |
| IQ4_XS | 4.25 | 54.2 GB | 56.6 GB | Won't fit | — | 5.09 |
| Q3_K_M | 3.91 | 50.0 GB | 52.3 GB | Won't fit | — | 5.61 |
| IQ3_M | 3.70 | 47.4 GB | 49.7 GB | Won't fit | — | 6.03 |
| IQ3_XXS | 3.06 | 39.4 GB | 41.8 GB | Won't fit | — | 7.47 |
| Q2_K | 2.63 | 34.1 GB | 36.4 GB | Won't fit | — | 8.93 |
| IQ2_XXS | 2.06 | 27.0 GB | 29.3 GB | Won't fit | — | 12.6 |
| IQ1_M | 1.75 | 23.1 GB | 25.4 GB | Won't fit | — | 15.7 |