Can Yi 1.5 34B Chat run on the DGX Spark?
Yes. At 8K context the best fit is F16, using about 66.5 GB of the 109 GB usable on the DGX Spark. Estimated speed ceiling: ~4 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | Max tok/s (est., 8K) |
|---|---|---|
| 2-bit (Q2_K) • | 13.4 GB Fits | 21 |
| 3-bit (Q3_K_M) | 17.1 GB Fits | 16 |
| 4-bit (Q4_K_M) • | 20.7 GB Fits | 13 |
| 5-bit (Q5_K_M) • | 24.1 GB Fits | 11 |
| 6-bit (Q6_K) • | 27.7 GB Fits | 10 |
| 8-bit (Q8_0) • | 35.5 GB Fits | 7 |
| 16-bit (F16/BF16) | 65.5 GB Fits | 4 |
• real GGUF file size from lmstudio-community/Yi-1.5-34B-Chat-GGUF. Speed is a bandwidth ceiling (273 GB/s), not a benchmark. Methodology.
More
Yi 1.5 34B Chat on every GPU and Mac · Every model that runs on the DGX Spark · Other models that fit 64 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (01-ai/Yi-1.5-34B-Chat), NVIDIA specs. Data as of 2026-11-17.