Velaura AI Raises $110M, Valued Over $1B

Velaura AI has successfully closed a $110 million Series A funding round, propelling its valuation beyond the $1 billion mark. This significant capital infusion underscores the growing demand for ultra-efficient AI compute solutions, particularly as power consumption emerges as a critical bottleneck for widespread AI deployment. The company's strategic pivot from its origins in Bitcoin mining hardware to AI acceleration is now validated by this substantial backing from both existing and new investors.

The funding was led by Seligman Ventures, with participation from Capricorn Investment Group, alongside returning investors including Mayfield, Maverick Silicon, Premji Invest, Samsung Catalyst Fund, and StepStone Group. Notably, MARA Holdings, a public Bitcoin miner that was previously a customer, also participated, signaling strong confidence in the company's technological evolution. This round effectively rebrands and recapitalizes the entity formerly known as Auradine, which had previously raised over $300 million for its Bitcoin mining ASIC development.

At the heart of Velaura AI's offering is its Titan Core technology, a digital chip intellectual property and design platform. The company asserts that Titan Core delivers a 2-4x improvement in performance per watt for core AI accelerator computations, maintaining existing performance levels. A key differentiator for Velaura AI is the proven commercial-scale deployment of its underlying architecture, having shipped in over 30 million ASICs with exceptional yield and reliability. This track record in high-volume manufacturing provides a tangible advantage over many AI chip startups relying solely on prototype benchmarks.

Velaura AI is targeting two primary markets with its power-efficient compute architecture: large-scale data center AI accelerators and Physical AI applications. The latter encompasses intelligent systems operating in the physical world, such as robots and drones, where on-device compute efficiency is paramount due to battery and thermal constraints. This dual focus addresses the escalating need for compute power across diverse AI applications, from complex reasoning models to embodied intelligence.

The company's leadership team brings a wealth of experience in semiconductor design and successful exits. Co-founder and CEO Rajiv Khemani previously co-founded Innovium, a high-bandwidth Ethernet switching ASIC company acquired by Marvell for approximately $1.1 billion. Chief Development Officer Manu Gulati is a renowned silicon architect with prior roles at Apple, Google, and Qualcomm, and co-founded NUVIA, which was acquired by Qualcomm for around $1.5 billion. The broader engineering team comprises talent from industry leaders like NVIDIA and Qualcomm.

The broader AI chip sector has seen significant investment, with startups attracting billions this year. However, Velaura AI differentiates itself by focusing on providing the foundational IP for accelerators rather than competing directly with established players like NVIDIA in the finished chip market. This IP-centric model allows Velaura AI to potentially serve as a partner to other AI hardware developers. The company is actively engaged in multiple hyperscaler collaborations, aiming to integrate its power-efficient compute solutions into next-generation AI infrastructure.

The substantial energy demands of AI are increasingly recognized as a critical infrastructure challenge. Velaura AI's emphasis on extreme power efficiency at the silicon level directly addresses this growing concern. As Dipender Saluja of Capricorn Investment Group noted, the company is tackling the problem at its core. This strategic positioning, combined with a proven engineering foundation and strong investor support, places Velaura AI to capitalize on the next wave of AI innovation, where compute economics and energy efficiency will be paramount.