local intelligence

can it run

Can the CPU-only (DDR5 tuned / 4-channel) run llama3.2 3B?

FULL

2.4 GB needed of 30 GB usable — headroom for context

~30GB usable · CPU (DDR5) · model file 2.0GB · 3B

~40–63 tok/s

computed roofline: 179.2 GB/s × 0.45–0.70 efficiency window / 2.0GB (Q4_K_M) — bands, never points. Small models (under ~4.5B) run BELOW these bands — kernel-launch overhead keeps them from saturating bandwidth (measured against community medians, 09-06).

Run llama3.2 3B on instead

Raspberry Pi 5 + AI Kit (Hailo-8L) — cheapest catalog machine that runs llama3.2 3B fully ($110)

every model the CPU-only (DDR5 tuned / 4-channel) can run — full list

explore the full catalog84 machines indexed · live prices · what each one can run
2026-09-17 · ← all models on the CPU-only (DDR5 tuned / 4-channel) · fit = weights + 4k KV vs usable memory