£300 Snapdragon mini PC runs an 80B model at 17 tok/s on CPU via ARM-ported Janas
WorldlyAd7946 · reddit · 2026-10-16
The author tested a £299.99 Lenovo IdeaCentre Mini (Snapdragon X1-26-100, 32GB, 35W, CPU only). With llama.cpp, Qwen3.6 35B A3B coder reads 118 and writes 24.5 tok/s (30 with MTP). They ported the x86-only Janas expert-cache/NVMe-streaming runtime to ARM (NEON kernels, i8mm, bit-exact, PRs coming), which runs Qwen3-Next 80B A3B — unloadable in llama.cpp at 32GB — at 17-18 tok/s with MTP. A 12GB expert cache still yields 24 tok/s on the 35B, suggesting 16GB Snapdragon devices could run it. USB4-linking two boxes already ran a 122B Qwen. Possibly the cheapest path to 32GB LPDDR5 today.
More from Infra
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Jensen Huang lays out NVIDIA's full Vera Rubin agentic system: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- ONNX to Transformers.js: running models directly in your browser — JosephJacks_ · 2026-10-23
- MiniMax M3.1 Flash free for 3 days on 0G, with 1M context and 512K output — MiniMax_AI · 2026-10-23
- "We bet on Cloudflare and it paid off": R2 and D1 at billion scale — dinasaur_404 · 2026-10-23
- Public data is drying up; private training data is turning into a browsable catalog — SucceededMind · 2026-10-23