[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-mobile-llms-are-wasting-nearly-half-their-battery":10,"sections":35},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":30,"feedback":34,"feedback_at":22,"cost_usd":34,"total_tokens":34},4327,"mobile-llms-are-wasting-nearly-half-their-battery","Mobile LLMs Are Wasting Nearly Half Their Battery","A new benchmarking study finds that misconfigured NPUs waste up to 40% of energy, and fixing that could cut mobile LLM power use by 54%.","Running AI models on your phone burns more battery than it should — and the culprit is mostly bad configuration, not bad hardware.\n\nResearchers have published what they call the first cross-layer measurement study of mobile LLM inference, testing five frameworks — including llama.cpp and GENIE — across CPU, GPU, and NPU backends. To do it, they built PowerBench, a profiling tool that tracks energy use per backend rather than just at the device level. The findings are blunt: NPUs show up to a 10x performance gap between frameworks, driven by how each handles custom operators and quantization. Worse, up to 40% of energy is wasted through scheduling mistakes alone — things like misconfigured thread counts and uncoordinated NPU sleep cycles.\n\nThe study also surfaces a counterintuitive split: NPUs are better at the \"prefilling\" phase of inference, where workloads are large and fixed-shape, but CPUs actually outperform NPUs during \"decoding,\" where requests are small and dynamic. That means the hardware most people assume is doing the heavy lifting is often the wrong tool for the job. Fixing the configuration, the researchers estimate, could cut energy consumption by up to 54.8% on NPU backends.\n\nMobile AI is being sold on privacy and low latency, but if the software stack is burning half the battery through avoidable scheduling errors, that pitch gets harder to make — especially as on-device models grow larger.","[\"mobile ai\",\"inference\",\"hardware\",\"energy efficiency\"]","2026-07-08T04:00:00.000Z","2026-07-08T05:56:16.685Z","2026-07-08T05:56:19.655Z","published",null,[],"ai",[26,27,28,29],"mobile ai","inference","hardware","energy efficiency",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.05475",0,{"sections":36},[37,41,46,51,56,60,65,70,75,80,85,89,94,99],{"name":38,"slug":24,"count":39,"latest_published_at":40},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":42,"slug":43,"count":44,"latest_published_at":45},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":47,"slug":48,"count":49,"latest_published_at":50},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":52,"slug":53,"count":54,"latest_published_at":55},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":57,"slug":28,"count":58,"latest_published_at":59},"Hardware",122,"2026-07-14T19:46:26.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":86,"slug":87,"count":83,"latest_published_at":88},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]