⬢github Python · 778 ★ +148 since we first saw it · pushed 1 d ago · MIT
StayLameBro/backburner
Your iPhone helps your Mac run a 27B model: faster prompt reading and more context over a USB-C cable
Backburner is a llama.cpp fork and iPhone app that lets a USB-C-connected iPhone help a MacBook run Qwen3.8-27B locally. The Mac and iPhone split transformer layers or share context attention across the cable, giving roughly 30-44% faster prompt processing and much larger context windows (up to ~140k-229k tokens) with identical outputs.
Why now: Recently created and quickly starred on GitHub's trending lists, it demos an unusual trick — using an idle iPhone as a second Apple-Silicon-class accelerator for local LLM inference — with published benchmarks.
Who it is for: Mac users with Apple Silicon and a recent iPhone or iPad who run large local models for coding agents and want faster prefill and more context without cloud services.
apple-siliconiosiphonellama-cppllm-inferencelocal-llmmacosmetalqwensme2
Stars over our 127 snapshots: 630 to 778, since 1 d ago.
Where people talked about it
- ⬢github new repos, most starred 12 min ago
- ⬢github new repos, most starred 3 h ago
- ⬢github new repos, most starred 1 d ago
API: https://socialmediatrends-api.osmike.com/v1/repos/StayLameBro/backburner