RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can Qwen3.5 27B run on the A100 40GB?

Yes. At 8K context the best fit is Q8_0, using about 28.0 GB of the 38.0 GB usable on the A100 40GB. Estimated speed ceiling: ~54 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K)10.8 GB Fits11.1 GB Fits12.6 GB Fits18.6 GB Fits149
3-bit (Q3_K_M) •13.7 GB Fits14.0 GB Fits15.5 GB Fits21.5 GB Fits115
4-bit (Q4_K_M) •16.7 GB Fits17.0 GB Fits18.5 GB Fits24.5 GB Fits93
5-bit (Q5_K_M) •19.4 GB Fits19.6 GB Fits21.1 GB Fits27.1 GB Fits79
6-bit (Q6_K) •22.0 GB Fits22.3 GB Fits23.8 GB Fits29.8 GB Fits69
8-bit (Q8_0) •27.8 GB Fits28.0 GB Fits29.5 GB Fits35.5 GB Tight54
16-bit (F16/BF16) •51.2 GB No51.5 GB No53.0 GB No59.0 GB No—

• real GGUF file size from unsloth/Qwen3.5-27B-GGUF. Speed is a bandwidth ceiling (1555 GB/s), not a benchmark. Methodology.

More

Qwen3.5 27B on every GPU and Mac · Every model that runs on the A100 40GB · Other models that fit 48 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (Qwen/Qwen3.5-27B), NVIDIA specs. Data as of 2026-11-17.