Electronics visibility eliminates energy waste at the chip level, and beyond

Noam Brousard, Vice President, Solutions Engineering, proteanTecs, explains how semiconductors are straining under GenAI workloads such as model training, inference, and high-performance AI applications, and are quietly bleeding energy across data centres. proteanTecs has invented ML-driven deep data health & performance monitoring, which enables AI chipmakers to reclaim 8 - 14% in power savings without sacrificing performance, by eliminating the energy waste baked into traditional design. This also decreases the amount of water needed to cool down servers. For a data centre like Meta’s Hyperion, that would mean savings of up to 700 megawatts or enough power for more than 200,000 homes.