MIT Lincoln Laboratory's Supercomputing Center has developed tools to reduce data center energy use by power-capping hardware and by improving the efficiency with which models are trained in some cases, their techniques reduce energy use by 80%.
Researchers from MIT CSAIL predict semantically meaningful and calibrated uncertainty intervals in the latent space of a generative adversarial network to know the exact limits in our ability to clean up, and reconstruct, partly blurred images.
Taking a magnifying glass to data center operations scienceblog.com - get the latest breaking news, showbiz & celebrity photos, sport news & rumours, viral videos and top stories from scienceblog.com Daily Mail and Mail on Sunday newspapers.