Tech news in 3 minutes
Omen AI’s plan to optimize data centers is all wet
AI-driven demand for compute power is causing data centers to seek more efficiency from GPU racks, leading to bacterial outbreaks in liquid cooling systems. The coolant, a mix of water and bacteria-inhibiting substance, can be adjusted to include more water for better heat absorption, but this increases contamination risk, clogging flows. Flushing the system can shut down a rack for five to six hours, costing millions. Omen AI offers a solution: a tiny spectrometer that monitors fluid health in real time, spotting bacterial growth before it becomes a problem. CEO Zach Laberge says it avoids risking downtime from lack of chemical insight. Today, Omen AI announced a $31 million Series A round led by Nava Ventures, with participation from CRV, Vanderbilt University, Mann+Hummel, Starhill Holdings, Hard Launch Capital, and executives from Bridgestone, GM, Johnson Controls, and TensorWave. Laberge, who founded his first company at age 14 in 2020 and later dropped out of high school, started Omen in 2024. The technology initially focused on fluid systems for heavy vehicles, replacing lab sample analysis with real-time detection of bacteria, copper, chromium (for pump wear), and silicon (for seal wear). Caterpillar dealerships were early customers, and Caterpillar also supplies gas turbines for data center power. About six months ago, dealerships asked Omen to monitor building-side systems, leading the company to focus on data centers, which have fluid in HVAC and chip cooling. Nava Ventures partner Cory Rellas noted Laberge's rare ability to earn respect from large corporations. Omen, having raised $40 million total, works with a dozen data center customers, including TensorWave, which builds an AMD-based AI compute cloud. TensorWave's president Piotr Tomasik said the fluid is a critical variable most of the industry flies blind on. Omen faces competition from Pyxis, which also launched data center coolant monitoring. Key advances include cheaper optical hardware and improved signal processing software, enabling real-time analytics at scale.