Apple unleashes a swarm of Mac Minions to share GPU resources for large AI tasks

What happened: Apple researchers described Minions, a distributed system that aggregates idle Apple Silicon GPUs across macOS devices to run large-scale machine learning inference workloads. Key numbers: Prior to adoption, some inference jobs had queue times exceeding ten hours; with Minions, tasks can often begin within minutes. Why it matters: Apple says the system makes GPU compute more readily available to its teams, and its hardware availability focus differs from Nvidia clusters, which target throughput and latency for large models.

Apple researchers described Minions, a distributed system that aggregates idle Apple Silicon GPUs across macOS devices to run large-scale machine learning inference workloads.

Apple says the system makes GPU compute more readily available to its teams, and its hardware availability focus differs from Nvidia clusters, which target throughput and latency for large models.