{"slug": "ask-hn-how-is-gpu-power-draw-measured-at-scale", "title": "Ask HN: How is GPU power draw measured at scale?", "summary": "A developer on Hacker News is asking how GPU power draw is measured at scale in self-hosted setups, sharing per-GPU data from NVML polling and questioning whether per-request attribution survives at rack scale or if monitoring shifts entirely to PDU/BMC level.", "body_md": "How do people measure power usage of GPUs at large (32x) self-hosted setups or small multi-rack setups? I've seen some PDUs which collect and transmit data, but I'm unsure of the processes and if/how people do this on small builds.\n\nCurrently, I collect NVML nvmlDeviceGetPowerUsage, polled at 100ms during inference, peak and mean per request, and get this type of data:\n\nmodel mean-power range (W) spread stdev\n\nqwen3-8b 114.3-121.9 7.6W 1.17\n\nllama-3.1-8b-instruct 104.7-122.1 17.4W 4.29\n\nqwen2.5-1.5b-instruct 53.7-73.0 19.3W 5.23\n\nmistral-7b-instruct-v0.3 96.2-120.0 23.8W 6.01\n\nqwen2.5-7b-instruct 88.7-124.5 35.8W 7.73\n\ngemma-3-1b-it 49.4-56.7 7.3W 2.13\n\nthis is per-GPU, single-card data - I don't know whether anything like per-request attribution survives at rack scale, or whether monitoring there happens entirely at the PDU/BMC level instead.\n\nComments URL: [https://news.ycombinator.com/item?id=48684938](https://news.ycombinator.com/item?id=48684938)\n\nPoints: 3\n\n# Comments: 0", "url": "https://wpnews.pro/news/ask-hn-how-is-gpu-power-draw-measured-at-scale", "canonical_source": "https://news.ycombinator.com/item?id=48684938", "published_at": "2026-06-26 10:35:10+00:00", "updated_at": "2026-06-26 11:05:58.584243+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-tools", "developer-tools"], "entities": ["NVML", "PDU", "BMC", "qwen3-8b", "llama-3.1-8b-instruct", "qwen2.5-1.5b-instruct", "mistral-7b-instruct-v0.3", "gemma-3-1b-it"], "alternates": {"html": "https://wpnews.pro/news/ask-hn-how-is-gpu-power-draw-measured-at-scale", "markdown": "https://wpnews.pro/news/ask-hn-how-is-gpu-power-draw-measured-at-scale.md", "text": "https://wpnews.pro/news/ask-hn-how-is-gpu-power-draw-measured-at-scale.txt", "jsonld": "https://wpnews.pro/news/ask-hn-how-is-gpu-power-draw-measured-at-scale.jsonld"}}