Can Yi 34B 200K run on the L40S?
Yes. At 8K context the best fit is Q8_0, using about 36.4 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~24 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | 15.0 GB Fits | 16.0 GB Fits | 21.6 GB Fits | 44.1 GB Tight | 59 |
| 3-bit (Q3_K_M) • | 17.0 GB Fits | 17.9 GB Fits | 23.5 GB Fits | 46.0 GB No | 52 |
| 4-bit (Q4_K_M) • | 20.7 GB Fits | 21.6 GB Fits | 27.3 GB Fits | 49.8 GB No | 42 |
| 5-bit (Q5_K_M) • | 24.1 GB Fits | 25.0 GB Fits | 30.7 GB Fits | 53.2 GB No | 36 |
| 6-bit (Q6_K) • | 27.7 GB Fits | 28.7 GB Fits | 34.3 GB Fits | 56.8 GB No | 31 |
| 8-bit (Q8_0) • | 35.5 GB Fits | 36.4 GB Fits | 42.1 GB Tight | 64.6 GB No | 24 |
| 16-bit (F16/BF16) | 65.5 GB No | 66.5 GB No | 72.1 GB No | 94.6 GB No | — |
• real GGUF file size from TheBloke/Yi-34B-200K-GGUF. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.
More
Yi 34B 200K on every GPU and Mac · Every model that runs on the L40S · Other models that fit 48 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (01-ai/Yi-34B-200K), NVIDIA specs. Data as of 2026-11-17.