{"rewrite":{"id":"r_e856ba861f1f2514b72d70f8","clusterId":"c_a91f4af348c6c99f03c82d94","slug":"apple-builds-virtual-gpu-cloud-from-idle-macs","model":"deepseek-v4-1-flash","headline":"Apple Builds Virtual GPU Cloud From Idle Macs","summary":"Apple's research team has developed Minions, a distributed system that gathers idle GPU capacity from Macs across the company and runs AI jobs on them as a virtual GPU cloud. Large jobs are split into small tasks, each handled by a Minion Executor on an Apple Silicon Mac, with results saved to Amazon S3. Apple says waits that once exceeded 10 hours now begin within minutes.","whyItMatters":"Apple cut internal AI queue times by treating hardware it already owns as a pool rather than buying more GPU servers, which is the cheaper path for any company sitting on large fleets of idle machines.","webCardHtml":"\u003cp\u003eThe system splits a large job into small tasks and hands each one to a Mac on the network. A program called Minion Executor pulls a task and runs it on the Apple Silicon GPU, then writes the result to Amazon S3. Each Mac is not carrying part of one model. Each one runs its own task independently.\u003c/p\u003e\u003cp\u003eBecause those Macs belong to people, they can sleep, drop off the network, or be reclaimed by their owners. Each machine sends a heartbeat every five minutes, and Minion Watchdog returns unfinished tasks from silent Macs to the queue for another machine to redo. Owners can also set which teams get priority on their GPU.\u003c/p\u003e","blueskyPost":"Minions runs on Macs already bought and sitting idle, so Apple's AI capacity story is about utilization, not new hardware. Results land in Amazon S3 rather than Apple infrastructure.","twitterPost":"Minions repurposes Macs Apple already bought, so the gain is utilization, not new silicon. Results land in Amazon S3.","threadsPost":"Minions turns Macs Apple already owns into GPU capacity, which makes this a utilization story more than a hardware story. The results get written to Amazon S3, so the compute may be Apple's but the storage layer is not. That split is the part worth watching.","newsletterBlurb":"Apple's research team has a distributed system called Minions that pools GPU capacity from idle Macs inside the company and runs AI processing on them. Jobs are split into small tasks, each executed on an Apple Silicon GPU, with heartbeats and a watchdog to recover tasks from machines that drop off. Apple says waits that once exceeded 10 hours now begin within minutes.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20261015-apple-minions/\",\"title\":\"Apple Develops 'Minions,' Building a Virtual GPU Cloud from Unused Macs, Cutting AI Processing Wait Times from Over 10 Hours to Minutes\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":4615,"outputTokens":593,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1792036339,"createdAt":"2026-10-15T03:49:21.000Z","publishedAt":"2026-10-15T03:51:08.000Z","updatedAt":"2026-10-15T03:51:08.000Z"},"cluster":{"id":"c_a91f4af348c6c99f03c82d94","canonicalTitle":"Appleが使われていないMacを集めて仮想GPUクラウドを構築する「Minions」を開発、AI処理の待ち時間を10時間以上から数分に短縮","representativeArticleId":"a_d5f86b1b6cf50b50ace903a6","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-10-15T01:57:00.000Z","lastSeenAt":"2026-10-15T01:57:00.000Z","updatedAt":"2026-10-15T03:51:06.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20261015-apple-minions/","title":"Appleが使われていないMacを集めて仮想GPUクラウドを構築する「Minions」を開発、AI処理の待ち時間を10時間以上から数分に短縮"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":null}