Apple researchers release 'Minions,' which pools idle Mac GPUs for distributed ML inference (community reports)
Verified (fact)
According to GeekNews, Apple researchers released Minions, a distributed system that splits large machine learning inference jobs across the idle Apple Silicon GPUs of multiple Macs. It can be used for tasks such as generating embeddings with Hugging Face models and running LLM inference with Ollama.
Needs fact-check (claim)
It is described as reducing the need to buy extra GPUs by using existing hardware and as supplementing limited central GPU infrastructure.
Why it matters (interpretation)
Using personal devices scattered across a company as inference resources is one way to cope with GPU shortages.
Confidence medium · official source pending
Sources
- Community Minions - 유휴 Mac의 GPU를 모아 머신러닝 추론 작업을 분산하는 Apple의 시스템 GeekNews
More about Apple
-
U.S. court orders OpenAI to hand over interview records for ex-Apple hires in Apple trade secrets suit (report)
A U.S. federal judge ruled that Apple can access OpenAI's interview records for former Apple employees. Private messages and testimony...
-
Apple unveils fourth-generation Apple TV 4K with Siri AI and far-field microphones (community reports)
According to GeekNews, Apple announced a fourth-generation Apple TV 4K with Siri AI and far-field microphones. Users can find content...
-
Apple unveils HomeView, its first home smart display with Siri AI and smart home control (report)
Apple unveiled 'HomeView', its first home smart display, in tabletop and wall-mounted versions. It combines Siri AI, music and video...
-
Apple discloses deal to hire team from personalized podcast startup Huxe and license its technology (report)
According to TechCrunch, Apple disclosed a deal to bring on the team of Huxe, a personalized podcast startup, and license its...