New Startup Niv-AI Tackles AI Power Use with Innovative Solutions

In a world where electricity is vital for artificial intelligence (AI), data centers struggle to manage power usage effectively, potentially leading to a 30% throttle in operations. Nvidia CEO Jensen Huang highlighted this inefficiency during the company's GTC customer conference, emphasizing the revenue lost with each wasted watt. Enter Niv-AI, a Tel Aviv-based startup, which has just emerged with $12 million in seed funding to address this issue by developing tools to optimize electricity usage in data centers. - Founders: CEO Tomer Timor and CTO Edward Kizis. - Backers include: Glilot Capital, Grove Ventures, Arc VC, Encoded VC, Leap Forward, and Aurora Capital Partners. - Niv-AI is developing rack-level sensors to measure power usage on GPUs at a millisecond level. The aim is to create an "intelligence layer" that better synchronizes power loads and enables data centers to utilize more of their existing capacity. - An operational system is anticipated in select US data centers over the next six to eight months. - Further challenges include managing surges in power demand as frontier labs run thousands of GPUs simultaneously, prompting costs for temporary energy storage or decreased GPU usage. This innovative approach attempts to alleviate these issues while easing the burden on the electrical grid and offering a more sustainable power consumption model.