Niv-AI exits stealth to wring more power performance out of GPUs

Niv-AI, a Tel Aviv-based startup founded in 2025, has emerged from stealth with $12 million in seed funding to address inefficiencies in GPU power usage in AI data centers. These centers often experience millisecond-scale power surges as GPUs switch tasks, leading operators to throttle GPU performance by up to 30% or invest in costly temporary energy storage. This results in significant wasted electricity and lost revenue, as highlighted by Nvidia CEO Jensen Huang. Niv-AI aims to solve this by deploying rack-level sensors that measure GPU power consumption with high precision and developing tools to optimize power management, thereby unlocking more of the existing GPU capacity. The company plans to use the collected data to build AI models that predict and synchronize power loads across data centers, effectively acting as a "copilot" for data center engineers. Niv-AI expects to have operational systems in several U.S. data centers within six to eight months. The startup envisions its technology as an "intelligence layer