TechnologyPhoto: HT BusinessApple's Minions Uses Idle Macs to Power AI Workloads
Apple researchers have created Minions, a system that uses idle Macs with M-series silicon as a virtual GPU cloud for machine learning inference. This approach reduces GPU queue times from up to 10 hours to nearly zero by pooling macOS devices' GPUs. The system includes a Minion Server, Executor, and Watchdog to manage and distribute AI compute tasks efficiently across multiple devices.
HT Business·


