SANTA CLARA, California — The artificial-intelligence chip race may no longer be just about building faster processors. Increasingly, the battle is about something far more basic: how much electricity those chips consume.
That shift has helped propel Velaura AI into unicorn territory.
The Silicon Valley chip-design startup has raised $110 million in a Series A funding round, pushing its valuation above $1 billion as investors place a major bet on technology designed to slash the enormous power requirements of AI computing infrastructure.
The round was led by Seligman Ventures, with Capricorn Investment Group joining as a new investor. Existing backers including Samsung Catalyst Fund, StepStone Group and Maverick Silicon also participated.
The fresh capital will be used to accelerate development and deployment of Velaura’s AI products while expanding its engineering and customer-facing teams.
But the bigger story behind the billion-dollar valuation is the problem Velaura is trying to solve.
AI’s Next Bottleneck Could Be Electricity
Generative AI has triggered an unprecedented expansion of data centers packed with power-hungry GPUs and other accelerators.
As companies race to deploy increasingly powerful models, electricity availability, cooling capacity and infrastructure costs are becoming critical constraints.
Velaura is positioning itself directly in the middle of that problem.
The company develops low-power silicon and software technologies for AI data centers as well as so-called physical AI applications, including robotics and autonomous systems.
Its flagship technology is Titan Core, a silicon-design and intellectual-property platform unveiled earlier in 2026.
Velaura says Titan Core can enable up to two times lower overall chip power consumption for AI accelerators and potentially save as much as 500 watts on a typical 1,000-watt GPU or XPU.
The underlying technology targets the energy consumed by matrix-multiplication operations—the mathematical calculations at the heart of AI training and inference.
According to earlier reporting by GamesBeat, Velaura says Titan Core can reduce the energy required for those operations by roughly two to four times through proprietary circuit and library technology.
Those are company performance claims and will ultimately depend on how the technology performs when integrated into commercial AI processors at scale.
Why Even Small Power Savings Could Be Worth Billions
A reduction of hundreds of watts per accelerator may sound relatively modest until it is multiplied across enormous AI clusters containing tens or hundreds of thousands of processors.
Electricity is only part of the equation.
More power consumption also means more heat, which requires additional cooling infrastructure. Data-center operators must therefore consider electricity supply, cooling systems, transformers and other infrastructure alongside the cost of the processors themselves.
Velaura argues that substantially lowering chip power consumption could allow operators to deploy more computing capacity within the same electrical envelope.
GamesBeat reported in March that the company estimated its technology could produce roughly $1,300 in electricity savings over three years per XPU, before additional cooling and infrastructure savings are considered.
At hyperscale, those economics become potentially significant.
Velaura’s Technology Is Already Drawing Hyperscaler Interest
Velaura has said it is working with leading hyperscaler partners on advanced semiconductor designs.
Earlier this year, the company said it had ongoing engagements involving 3-nanometer and 2-nanometer process nodes and was discussing integrating Titan Core technology into next-generation AI accelerators.
The company also says the technology behind Titan Core has been validated through more than 30 million advanced ASICs deployed in production over several years.
That experience could prove important in a semiconductor industry where moving an innovative design from laboratory demonstrations into high-volume manufacturing is notoriously difficult.
From Auradine to a Billion-Dollar AI Bet
Velaura AI wasn’t always focused so heavily on artificial intelligence.
The company was previously known as Auradine and built expertise in highly efficient custom silicon, including chips used for Bitcoin mining.
It subsequently rebranded as Velaura AI as it increased its focus on ultra-low-power AI computing.
GamesBeat reported earlier this year that Velaura had about 85 employees at the time and was headquartered in Santa Clara, California. The company had previously raised roughly $300 million across its earlier businesses before the latest financing.
Its leadership also carries considerable semiconductor experience.
CEO and co-founder Rajiv Khemani has positioned power efficiency as a defining challenge for the next generation of artificial intelligence.
“The next era of AI will be defined not only by better models, but also by fundamentally better compute economics,” Khemani said following the latest financing.
Investors Are Chasing the Infrastructure Behind the AI Boom
Velaura’s billion-dollar valuation also reflects a broader change in where investors see opportunity in artificial intelligence.
The first phase of the generative-AI boom concentrated enormous attention on models and applications.
The next phase is increasingly focused on the infrastructure required to keep those models running: processors, networking, memory, cooling, electricity generation and massive data-center campuses.
That creates opportunities for companies capable of reducing the cost of AI computation rather than merely increasing raw processing performance.
Velaura’s pitch is straightforward: if AI companies cannot easily obtain more electricity, they will need to extract substantially more computing power from every available watt.
That is a potentially enormous market—but it is also fiercely competitive.
Velaura will be competing for relevance in an ecosystem dominated by established semiconductor companies and a rapidly growing collection of heavily funded AI-chip startups.
Crossing the $1 billion valuation threshold therefore represents investor confidence, not proof that Titan Core will become an industry standard.
The real test comes next: whether Velaura can turn its promised power savings into technology that the world’s largest chip designers and data-center operators deploy at massive scale.
And in an AI industry increasingly constrained not just by chips but by the electricity needed to run them, solving that problem could ultimately be worth far more than Velaura’s new billion-dollar valuation.
Leave a Reply