{"slug": "researchers-turned-used-pixel-phones-into-a-data-center", "title": "Researchers Turned Used Pixel Phones Into a Data Center", "summary": "Google Research and UC San Diego have repurposed 2,000 retired Pixel phones into a data center that matches the performance of 50 traditional servers, achieving higher per-thread CPU scores than AMD EPYC server cores in benchmarks. The cluster, set to launch in Fall 2026, reduces carbon emissions by reusing hardware instead of manufacturing new servers, though it is best suited for budget-constrained institutions running parallelizable workloads.", "body_md": "That drawer full of “obsolete” phones just became more valuable than you thought. [Google](https://www.gadgetreview.com/german-court-holds-google-liable-for-false-ai-overview-claims) Research and UC San Diego are proving that [retired Pixel phones](https://research.google/blog/a-low-carbon-computing-platform-from-your-retired-phones/) can match—and sometimes beat—server-grade hardware in head-to-head performance tests, turning [ 2,000 discarded devices](https://www.tomshardware.com/desktops/servers/researchers-recycle-old-phones-and-cluster-them-into-computing-platforms-says-processors-on-modern-smartphones-deliver-higher-single-core-performance-than-comparable-multicore-servers) into a data center that delivers the computing power of\n\n**50** traditional servers.\n\n## The Performance Reality Check\n\nThe numbers flip conventional wisdom on its head. A **2023 Pixel Fold’s** performance cores score higher than [ AMD EPYC](https://www.amd.com/en/products/processors/server/epyc.html) server cores in SPEC CPU 2017 benchmarks when compared per-thread. Google’s testing shows that clustering\n\n**25 to 50** phone motherboards delivers the same CPU throughput as a modern dual-socket server.\n\n[In classroom trials](https://www.techspot.com/news/112762-researchers-turning-old-pixel-phones-data-center-they.html), a 20-phone micro-cluster processed assignments from **75 students** in parallel programming courses—finishing the job in **50 seconds**, faster than comparable AWS instances.\n\n## From Android to Data Center\n\nThe transformation is surgical: remove screen, battery, cameras, and speakers, leaving just the motherboard with its system-on-chip, RAM, and storage. These boards get mounted in server racks, powered by centralized supplies, and networked like traditional nodes.\n\nAndroid gets replaced with standard [ Linux distributions](https://hothardware.com/news/google-turns-thousands-of-pixel-phones-into-a-low-carbon-data-center), then managed through\n\n**Kubernetes**—making a cluster of phones look identical to any other cloud infrastructure to users and applications.\n\n## The Carbon Angle That Actually Matters\n\nHere’s where the sustainability math gets interesting: roughly **50%** of a smartphone’s manufacturing emissions come from the motherboard and processor assembly. With typical upgrade cycles of **three to four years**, functional computing power gets discarded while companies build new servers for the same workloads.\n\nThe UCSD cluster, launching [ Fall 2026](https://www.kucoin.com/news/flash/google-and-ucsd-to-build-low-carbon-cloud-using-2000-retired-pixel-phones), aims to keep hardware productive longer rather than manufacturing fresh silicon for student coursework and research projects.\n\n## Reality Check on the Limits\n\nThis isn’t replacing Google’s own data centers anytime soon. Managing thousands of heterogeneous phone boards introduces operational complexity that runs counter to typical data center standardization. Questions remain about long-term reliability under continuous server duty—phones weren’t designed for **24/7** rack life.\n\nThe sweet spot appears to be [budget](https://www.gadgetreview.com/the-trump-phone-a-budget-phone-priced-like-flagship-gold)-constrained institutions running parallelizable workloads like course autograding and batch analytics, where getting compute capacity at “a fraction of the usual cost” matters more than maximum reliability guarantees.\n\nThe 2,000-phone cluster represents something bigger than clever recycling. It challenges how we think about meeting growing compute demand—not just by [building more data centers](https://www.gadgetreview.com/openai-and-partners-launch-500-billion-stargate-project), but by aggregating the billions of capable devices already in circulation.", "url": "https://wpnews.pro/news/researchers-turned-used-pixel-phones-into-a-data-center", "canonical_source": "https://www.gadgetreview.com/researchers-turned-used-pixel-phones-into-a-data-center", "published_at": "2026-06-15 17:28:06+00:00", "updated_at": "2026-06-15 17:42:42.573110+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-research"], "entities": ["Google", "UC San Diego", "Pixel", "AMD EPYC", "Kubernetes", "AWS"], "alternates": {"html": "https://wpnews.pro/news/researchers-turned-used-pixel-phones-into-a-data-center", "markdown": "https://wpnews.pro/news/researchers-turned-used-pixel-phones-into-a-data-center.md", "text": "https://wpnews.pro/news/researchers-turned-used-pixel-phones-into-a-data-center.txt", "jsonld": "https://wpnews.pro/news/researchers-turned-used-pixel-phones-into-a-data-center.jsonld"}}