MIT's Delimitrou uses ML to boost data centers running at ~15% capacity
moniquejmorrow · x · 2026-10-30
MIT associate professor Christina Delimitrou's team applies machine learning to redesign software systems to fit existing hardware, making cloud computing more energy-efficient, predictable, and secure.
A striking detail: many large computing systems run at only 15% capacity. As AI demand grows and data center energy use strains grids, her work asks whether we're making good use of the infrastructure we already have.
Related event: MIT Professor Uses Machine Learning to Boost Data Center Efficiency(2 posts)→
More from Infra
- Dev blasts RunPod: on-demand B200 compute 'unreliable at best' for consistent workloads — Kyrannio · 2026-10-30
- 'Girls only want one thing': an $496,639 8x NVIDIA B300 deep learning server — Kyrannio · 2026-10-30
- SAM3 now runs offline on your GPU: ONNX export hits 0.37s on M5, 0.08s on RTX 3080 Ti — CSProfKGD · 2026-10-30
- Same model, different hardware: dev finds A100s vs B300s produce different outputs — Kyrannio · 2026-10-30
- Small devs can't reliably rent compute: grabbing 8 B300s on Modal is 'a shot in the dark' — Kyrannio · 2026-10-30
- SanDisk: ~50% of FY27 NAND supply locked in multi-year deals, datacenter bit demand +40% in 2027 — tengyanAI · 2026-10-30