AI Weekly Malaysia

Back to items Summaries

Show HN: Open-source engine running Gemma 4 26B in 2 GB RAM on any M-series Mac

ID
9450
Status
summarized
Published
29 Jul 2026, 11:05 PM
Fetched
31 Jul 2026, 11:59 PM
Provider
Hacker News
Category
dev-community
Original URL
https://github.com/drumih/turbo-fieldfare
Source URL
https://hnrss.org/best

Summary

Score
7.5
Created
01 Aug 2026, 12:01 AM
Tags
Audience
developersai_ml_learnersvibe_coders

What happened

TurboFieldfare is an open-source Swift + Metal inference engine that runs Gemma 4 26B-A4B in ~2 GB of RAM on any Apple Silicon Mac, including 8 GB models. It keeps only the 1.35 GB shared core and FP16 KV cache in memory, streaming individual experts from SSD per token instead of loading the full 14.3 GB model. The project includes 103 measured benchmark results across kernels, caching, I/O, prefill, and decode.

Why it matters

If you develop on an M-series Mac with limited RAM, this lets you run a 26B-parameter model locally without cloud API costs or memory upgrades. The expert-streaming-from-SSD approach is worth studying if you build on-device inference tooling, as it demonstrates a concrete memory budgeting strategy that trades I/O bandwidth for RAM footprint.

Discussion angle

Compare the expert-streaming-from-SSD tradeoff against quantized models or cloud APIs for a Malaysian developer on an 8 GB M1 MacBook Air — when does the SSD I/O latency make this impractical versus paying per-token API costs?

Top