Velaura AI Raises USD 110M Series A at USD 1B Valuation to Build Power-Efficient AI Chips

Velaura AI, a Santa Clara-based semiconductor company focused on ultra-low-power AI compute infrastructure, raised USD 110 million in Series A financing on August 18, pushing its valuation above USD 1 billion. Seligman Ventures led the round, with new participation from Capricorn Investment Group and Prosperity7 Ventures, and continued backing from existing investors including Mayfield, Maverick Silicon, MARA, Premji Invest, Samsung Catalyst Fund, and StepStone Group.

The Technology

Velaura's core product is the Titan Core silicon platform, designed to deliver two to four times better performance per watt than competing architectures for mathematical operations at the heart of AI inference and training. The company focuses on the intersection of ultra-low-power silicon design and the software stack required to make that hardware usable at production scale.

The Team

Velaura's founding team and engineering leadership have backgrounds at Apple, NVIDIA, Google, Qualcomm, and Marvell. Several founders have previously shipped semiconductor products at scale, which matters in a market where the gap between an impressive chip demo and commercially deployable silicon is where most companies fail.

Market Applications

Velaura's technology targets AI data centres where power consumption is the primary operating cost constraint, as well as robotics, drones, and physical AI systems where power efficiency determines whether a device is commercially viable on battery. The company is commercialising across both markets from its current funding base.

What This Means for Founders

AI founders who are building applications that need to run efficiently at the edge, in mobile devices, or in power-constrained environments should watch the semiconductor layer closely. Velaura's raise signals that institutional investors believe the next wave of AI hardware differentiation will be won on efficiency rather than raw compute. For application-layer founders, better and cheaper inference chips mean lower infrastructure costs, which improves unit economics across the board.

Find investors backing the full AI stack on AngelLinx @ angellinx.ai/register.

AngelLinx Intelligence | angellinx.ai