Can falcon 40b run on the RTX 5070?
No. Even at 2-bit with 8K context, falcon 40b needs about 16.1 GB, more than the 11.4 GB usable on the RTX 5070.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | 15.7 GB No | 16.1 GB No | 19.0 GB No | 30.2 GB No | — |
| 3-bit (Q3_K_M) • | 19.7 GB No | 20.2 GB No | 23.0 GB No | 34.2 GB No | — |
| 4-bit (Q4_K_M) • | 24.7 GB No | 25.2 GB No | 28.0 GB No | 39.3 GB No | — |
| 5-bit (Q5_K_M) • | 29.5 GB No | 30.0 GB No | 32.8 GB No | 44.1 GB No | — |
| 6-bit (Q6_K) • | 33.1 GB No | 33.6 GB No | 36.4 GB No | 47.6 GB No | — |
| 8-bit (Q8_0) | 42.4 GB No | 42.9 GB No | 45.7 GB No | 56.9 GB No | — |
| 16-bit (F16/BF16) | 78.9 GB No | 79.4 GB No | 82.2 GB No | 93.4 GB No | — |
• real GGUF file size from mradermacher/falcon-40b-i1-GGUF. Speed is a bandwidth ceiling (672 GB/s), not a benchmark. Methodology.
Hardware that does fit it
RX 7900 XTRTX 3090RTX 3090 TiRTX 4090RX 7900 XTXRTX 5090
More
falcon 40b on every GPU and Mac · Every model that runs on the RTX 5070 · Other models that fit 12 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (tiiuae/falcon-40b), NVIDIA specs. Data as of 2026-11-17.