Can deepseek moe 16b base run on the A100 80GB?
Yes. At 8K context the best fit is F16, using about 32.8 GB of the 76.0 GB usable on the A100 80GB. Estimated speed ceiling: ~360 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | Max tok/s (est., 8K) |
|---|---|---|
| 2-bit (Q2_K) • | 7.7 GB Fits | 1760 |
| 3-bit (Q3_K_M) • | 9.4 GB Fits | 1392 |
| 4-bit (Q4_K_M) • | 11.6 GB Fits | 1088 |
| 5-bit (Q5_K_M) • | 13.0 GB Fits | 953 |
| 6-bit (Q6_K) • | 15.1 GB Fits | 805 |
| 8-bit (Q8_0) • | 17.7 GB Fits | 678 |
| 16-bit (F16/BF16) | 32.0 GB Fits | 360 |
• real GGUF file size from mradermacher/deepseek-moe-16b-base-GGUF. Speed is a bandwidth ceiling (2039 GB/s), not a benchmark. Methodology.
More
deepseek moe 16b base on every GPU and Mac · Every model that runs on the A100 80GB · Other models that fit 64 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (deepseek-ai/deepseek-moe-16b-base), NVIDIA specs. Data as of 2026-11-17.