https://www.youngju.dev/blog 천천히 올바르게. AI Researcher & DevOps Engineer Youngju's tech blog. GPU/CUDA, LLM, MLOps, Kubernetes AI workloads, distributed training, and data engineering. ko fjvbn2003@gmail.com (Youngju Kim) fjvbn2003@gmail.com (Youngju Kim) Tue, 16 Jun 2026 00:00:00 GMT https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling.en https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling.en As AI capex explodes and data centers grow to gigawatt scale, power and cooling have emerged as the dominant constraints. We map the big picture: surging rack power density, the limits of air cooling versus direct liquid and immersion cooling, PUE, grid constraints, and the perf/watt race among chips (Vera Rubin targeting roughly 10x). Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) datacenterai-infrastructurecoolingpowerliquid-coolingpuesustainability https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling.ja https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling.ja AI capex の急増とともにデータセンターがギガワット規模へと大きくなり、電力と冷却が最大の制約として浮上しました。ラック電力密度の急上昇、空冷の限界と直接水冷/液浸、PUE、電力網の制約、チップの perf/watt 競争(Vera Rubin の約 10 倍目標)まで、インフラの大きな絵を整理します。 Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) datacenterai-infrastructurecoolingpowerliquid-coolingpuesustainability https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-datacenter-power-cooling AI capex 급증과 함께 데이터센터가 기가와트 규모로 커지면서 전력과 냉각이 최대 제약으로 떠올랐습니다. 랙 전력 밀도 급상승, 공랭의 한계와 직접수냉/액침, PUE, 전력망 제약, 칩의 perf/watt 경쟁(Vera Rubin 약 10배 목표)까지 인프라의 큰 그림을 정리합니다. Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) datacenterai-infrastructurecoolingpowerliquid-coolingpuesustainability