▲ local intelligence

can it run

Can the Iris Xe 96EU (iGPU) run qwen3 32B?

NO

21.0 GB vs 13.5 GB usable — does not fit

~13.5GB usable · Iris Xe (iGPU) · model file 20GB · 32B

~1.3–4.1 tok/s

computed roofline: 102.4 GB/s × 0.25–0.80 efficiency window / 20.0GB (Q4_K_M) — bands, never points.

Run qwen3 32B on instead

→ Tesla V100 32GB (used) — cheapest catalog machine that runs qwen3 32B fully ($700)

→ every model the Iris Xe 96EU (iGPU) can run — full list

→explore the full catalog91 machines indexed · live prices · what each one can run
2026-10-01 · ← all models on the Iris Xe 96EU (iGPU) · fit = weights + 4k KV vs usable memory