Velaura AI Hits $1 Billion Valuation With $110M Series A to Slash AI's Power Bill

While most of the AI chip funding spotlight this year has chased companies trying to out-compute Nvidia, a Silicon Valley startup just raised nine figures by betting on the opposite proposition: that the next bottleneck in AI is not raw performance, but watts. On August 18, 2026, Velaura AI announced a $110 million Series A round that values the chip designer at more than $1 billion, with investors backing technology that can lower power consumption and operating costs at AI data centers.

The round was led by Seligman Ventures, with participation from new investor Capricorn Investment Group. Existing investors Samsung Catalyst Fund, StepStone Group and Maverick Silicon also joined — a syndicate that reads like a who’s-who of strategic semiconductor capital, anchored by Samsung’s venture arm.

What Velaura Actually Builds

Velaura AI develops low-power chips and software technologies for data centers and so-called “physical AI” applications — a category that includes robotics and autonomous systems. Founded in 2022 and based in Silicon Valley, the company operates under the leadership of co-founder and CEO Rajiv Khemani, alongside Chief Development Officer Manu Gulati and a leadership bench pulled from the upper ranks of the chip industry.

Its core product is Titan Core, a proprietary chip design and IP platform unveiled in March 2026 that targets what the company argues is the biggest problem in AI data center silicon: power. Velaura claims its technology can deliver up to 2x lower power consumption for AI accelerators — including GPUs and custom AI processors — by attacking efficiency at the design-IP level rather than building yet another general-purpose chip to compete head-on with established accelerators.

The distinction matters. Velaura is not primarily trying to sell racks of silicon. Instead, it licenses its power-saving intellectual property to companies that already build or deploy accelerators, wrapping the IP in a software stack that spans hyperscale data centers down to edge devices and robots.

An Arm-Style Business Model, Priced in Watts

The most distinctive detail in Reuters’ reporting is Velaura’s business model. The startup charges customers an upfront licensing fee plus a royalty tied to a share of the power savings they actually achieve — a structure CEO Rajiv Khemani confirmed is deliberately similar to Arm’s per-chip licensing model from the era before Arm began manufacturing its own chips.

It is an elegant alignment of incentives: the company only earns its royalty upside if customers genuinely burn less electricity. And electricity is precisely the line item threatening the economics of the entire AI buildout. As hyperscalers sign gigawatt-scale power purchase agreements and utilities delay interconnects, a technology that halves the power draw of an accelerator fleet converts directly into billions of dollars of operating cost — and into capacity headroom that is arguably scarcer than capital.

Khemani framed the thesis in a statement: “The next era of AI will be defined not only by better models, but also by fundamentally better compute economics.”

The traction claim behind the round is bold. Khemani told Reuters that Velaura is engaged with three of the four largest cloud computing providers, though he declined to name them. If even one of those engagements converts to a licensed deployment, the royalty stream could scale unusually fast for a company of this size.

From Auradine to Velaura: A Rebrand, A Reset

Velaura AI is less than two years old in its current form — but its cap table is not. The company was formerly known as Auradine, a blockchain-and-AI infrastructure startup that raised a $153 million Series C in 2025. In March 2026 it rebranded to Velaura AI to sharpen its focus on ultra-low-power AI compute, and unveiled Titan Core weeks later. Counting the prior incarnation, the company has now raised roughly $300 million to date, with investors including Celesta Capital among its earlier backers.

The “Series A” label on a nine-figure round for a company with that history signals something deliberate: a fresh corporate identity, a new product thesis, and a valuation reset — from a reported $1.05 billion — that gives new money a clean entry point into the power-efficiency story.

Why Now: Efficiency as the New Frontier

The timing of this round is not accidental. Three currents in the 2026 AI infrastructure market converge exactly where Velaura sits:

Power has become the binding constraint. The industry’s bottlenecks have migrated from chips (2023), to HBM memory (2024-25), to electricity and grid access (2026). Startups that reduce watts per token now address the scarcest resource in the stack.

Investors are seeking non-circular AI exposure. After a year of scrutiny over circular financing deals — Nvidia’s $105 billion guarantee arrangement with OpenAI being only the latest example — capital is hunting for infrastructure plays whose returns do not depend on a customer-investor loop. A licensing-and-royalty model with savings-linked pricing is about as far from circular as AI infrastructure gets.

Physical AI is opening a second market. Robotics and autonomous systems cannot carry data-center-class power budgets. Ultra-low-power compute is not a nice-to-have for robots — it is a precondition for the category. Velaura’s dual focus on data centers and physical AI positions it on both sides of the watt divide.

The Risks

Skeptics will note the challenges. The “up to 2x” claims are vendor numbers; independent validation at hyperscaler scale remains pending. Nvidia’s own roadmap relentlessly improves performance-per-watt every generation, and Arm itself is aggressively expanding deeper into AI data center IP — a much larger competitor playing the same licensing game. And the savings-linked royalty model, while elegant, requires auditable measurement of counterfactual power draw — a non-trivial commercial and technical negotiation.

But the direction of travel is clear. In a year when the industry’s biggest checks have gone to buying more of everything — more GPUs, more gigawatts, more racks — Velaura’s $1 billion valuation marks a bet that the smarter money is in needing less.