Niv-AI Emerges from Stealth to Optimize GPU Power Efficiency in AI Data Centers
In the rapidly evolving landscape of artificial intelligence, the demand for computational power has surged, placing unprecedented strain on data centers and their energy resources. Addressing this critical challenge, Tel Aviv-based startup Niv-AI has emerged from stealth mode, announcing a $12 million seed funding round aimed at enhancing GPU power efficiency.
Founded in 2025 by CEO Tomer Timor and CTO Edward Kizis, Niv-AI is backed by prominent investors including Glilot Capital, Grove Ventures, Arc VC, Encoded VC, Leap Forward, and Aurora Capital Partners. The company’s mission is to revolutionize how data centers manage the power consumption of GPUs, which are integral to AI processing.
The Power Dilemma in AI Data Centers
As AI models become more complex, the computational demands on data centers have intensified. GPUs, the workhorses of AI computations, experience rapid and unpredictable power surges during tasks such as model training and inference. These fluctuations can lead to inefficiencies, with data centers often resorting to throttling GPU performance or investing in costly energy storage solutions to manage these surges.
Nvidia CEO Jensen Huang highlighted this issue during a keynote at the company’s annual GTC conference, stating, There is so much power squandered in these AI factories. Every unused watt is revenue lost. This sentiment underscores the pressing need for more efficient power management solutions in AI data centers.
Niv-AI’s Innovative Approach
Niv-AI aims to tackle this challenge by deploying advanced sensors capable of measuring GPU power usage at millisecond intervals. By capturing detailed power consumption data, the company seeks to develop tools that enable data centers to manage energy use more effectively, thereby reducing waste and enhancing overall performance.
The startup’s strategy involves several key steps:
1. Data Collection: Implementing rack-level sensors to monitor real-time power usage across GPUs.
2. Analysis: Understanding the specific power profiles associated with various deep learning tasks.
3. Mitigation Techniques: Developing strategies to synchronize and predict power loads, allowing data centers to unlock additional capacity without compromising performance.
By leveraging the data collected, Niv-AI plans to train AI models capable of predicting and managing power consumption patterns. This copilot system for data center engineers aims to optimize energy use, reduce operational costs, and minimize environmental impact.
Industry Implications and Future Prospects
The introduction of Niv-AI’s technology comes at a pivotal moment for the AI industry. As data centers grapple with the dual challenges of escalating computational demands and energy efficiency, innovative solutions like those proposed by Niv-AI are poised to make a significant impact.
Lior Handelsman, a partner at Grove Ventures and a member of Niv-AI’s board, emphasized the necessity for change, stating, We just can’t continue building data centers the way we build them now. This perspective reflects a broader industry consensus on the need for sustainable and efficient infrastructure to support the next generation of AI applications.
Niv-AI’s roadmap includes deploying its system in select U.S. data centers within the next six to eight months. By providing precise power management tools, the company aims to help data centers maximize their existing resources, reduce reliance on supplementary energy storage, and avoid performance throttling.
The successful implementation of Niv-AI’s solutions could set a new standard for energy efficiency in AI data centers, offering a scalable model for managing the complex interplay between computational power and energy consumption.
Conclusion
As artificial intelligence continues to drive technological advancements, the infrastructure supporting these innovations must evolve to meet new challenges. Niv-AI’s emergence from stealth mode and its focus on optimizing GPU power efficiency represent a significant step toward more sustainable and efficient AI data centers. With substantial backing and a clear vision, Niv-AI is well-positioned to lead the charge in transforming how data centers manage energy in the age of AI.
