{"slug": "daikin-and-ntt-data-test-ai-data-center-cooling", "title": "Daikin And NTT DATA Test AI Data Center Cooling", "summary": "Daikin and NTT DATA Japan will conduct a proof of concept from July 2026 to March 2027 for an AI-driven data center cooling system at an NTT DATA facility in Japan. The system predicts server thermal conditions from power and temperature data to coordinate HVAC, chillers, and liquid cooling, aiming to improve energy efficiency and operational automation for dense AI racks.", "body_md": "# Daikin And NTT DATA Test AI Data Center Cooling\n\nDaikin and NTT DATA Japan will run a **July 2026** to **March 2027** proof of concept for AI-driven data center cooling at an NTT DATA facility in Japan. The companies say the system will predict internal server thermal conditions from indirect signals such as power draw and temperature, then coordinate **HVAC**, chillers, and **liquid-cooling** equipment. For AI infrastructure teams, the practical point is that dense AI racks are making cooling control part of the compute stack, not just a facilities problem. If the trial improves energy efficiency and operational automation, it gives operators a measured path to reduce overcooling without waiting for a full hardware refresh.\n\nAI infrastructure capacity increasingly depends on how well operators manage the physical layer around accelerators. This trial matters because it treats cooling as a dynamic control problem tied to server telemetry, not a fixed facilities setting that can be overprovisioned and forgotten.\n\n### What happened\n\nDaikin and NTT DATA Japan said they will begin a proof of concept in July 2026 for an AI-driven data center cooling optimization system. Daikin says the test will run through March 2027 at an NTT DATA data center in Japan and will evaluate energy efficiency, electricity costs, and operational automation.\n\n### Technical context\n\nThe proposed system predicts internal server thermal conditions from indirect data such as power consumption and temperature. It then coordinates air conditioning, chillers, and liquid-cooling equipment so cooling responds to changing heat loads from dense AI servers.\n\n### For practitioners\n\nThe useful signal is the integration pattern. AI infrastructure teams should expect thermal prediction, equipment control, and workload telemetry to become part of capacity planning, especially where accelerator density pushes conventional cooling assumptions.\n\n### What to watch\n\nThe key evidence will be measured savings, reliability under variable AI workloads, and whether the approach can be generalized across data centers without expensive site-specific tuning.\n\n## Key Points\n\n- 1Daikin and NTT DATA will test AI-driven cooling optimization at an NTT DATA Japan facility from July 2026.\n- 2The system predicts server thermal conditions from power and temperature signals, then coordinates HVAC, chillers, and liquid cooling.\n- 3AI infrastructure teams should watch whether measured energy savings justify integrating cooling control with dense accelerator operations.\n\n## Scoring Rationale\n\nThis is a solid infrastructure story because it connects AI workload growth with data center cooling automation, but it is still a planned proof of concept rather than a deployed commercial result. The impact is highest for operators managing dense AI racks and lower for general ML practitioners until measured efficiency results are published.\n\n## Sources\n\nPublic references used for this report.\n\nPractice interview problems based on real data\n\n1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.\n\n[Try 250 free problems](/problems)", "url": "https://wpnews.pro/news/daikin-and-ntt-data-test-ai-data-center-cooling", "canonical_source": "https://letsdatascience.com/news/daikin-and-ntt-data-test-ai-data-center-cooling-02fe26b1", "published_at": "2026-07-06 19:25:08+00:00", "updated_at": "2026-07-07 01:07:21.089714+00:00", "lang": "en", "topics": ["ai-infrastructure", "machine-learning", "ai-products"], "entities": ["Daikin", "NTT DATA", "NTT DATA Japan"], "alternates": {"html": "https://wpnews.pro/news/daikin-and-ntt-data-test-ai-data-center-cooling", "markdown": "https://wpnews.pro/news/daikin-and-ntt-data-test-ai-data-center-cooling.md", "text": "https://wpnews.pro/news/daikin-and-ntt-data-test-ai-data-center-cooling.txt", "jsonld": "https://wpnews.pro/news/daikin-and-ntt-data-test-ai-data-center-cooling.jsonld"}}