Santa Clara startup Velaura AI announced on August 18, 2026, that it secured $110 million in a Series A round that values the company north of $1 billion, betting that hyperscalers will pay handsomely for technology that promises to slash AI chip power consumption in half. Seligman Ventures led the financing announced Monday, with backing from Capricorn Investment Group, Mayfield, Maverick Silicon, MARA, Premji Invest, Samsung Catalyst Fund and StepStone Group.
The pitch arrives as cloud providers confront a mounting electricity crisis. Velaura's Titan Core platform isn't a GPU itself but rather silicon intellectual property that chip vendors and hyperscalers can integrate into their own AI accelerators. The company claims it can shave 500 watts off a typical 1,000-watt GPU while maintaining performance, targeting the power-intensive matrix-multiplication operations that drive modern machine learning.
Whether that promise holds at scale remains to be tested in production environments. Velaura said its technology has been validated "across tens of millions of ASICs" in company materials, though the specifics of those deployments weren't disclosed.
An Industry Choking on Kilowatts
Data centers are expected to consume 565 terawatt-hours of electricity in 2026, up 26 percent from the previous year, according to Gartner projections released in June. The International Energy Agency has forecast an even steeper trajectory: power demand from data centers could more than double to roughly 945 terawatt-hours by 2030, driven largely by AI training and inference workloads.
"AI's energy footprint is on track to become one of the most significant infrastructure challenges of this decade," said Dipender Saluja, managing partner at Capricorn Investment Group.
That presents an opening for companies like Velaura, which sells design libraries and low-voltage implementation services rather than finished chips. Customers get a drop-in optimization for their own silicon at advanced process nodes like 3-nanometer and 2-nanometer fabrication. Velaura says it delivers two-to-four-times improvements in performance-per-watt for mathematical operations. The company also markets the technology for what it calls "Physical AI" systems: robots, drones, and other embodied platforms where thermal constraints are even tighter than in data centers.
"The next era of AI will be defined not only by better models, but also by fundamentally better compute economics," Rajiv Khemani, Velaura's co-founder and CEO, said in a statement.
A Fresh Face on an Old Company
Velaura only recently adopted that name. The startup rebranded from Auradine on March 3, the same day it brought on Manu Gulati as co-founder and chief development officer. Gulati co-founded NUVIA, the chip startup Qualcomm acquired, and previously held engineering roles at Apple, Google, and Qualcomm. The company's executive roster includes veterans from NVIDIA, Marvell, and Palo Alto Networks. Velaura AI employed between 51 and 200 people according to LinkedIn as of August 19, 2026, with job postings spanning Santa Clara, Austin, Boston, Durham, and Bangalore.
The startup's LinkedIn page referenced a Reuters exclusive on the funding round, noting engagements with "three of the four largest cloud providers," though that article couldn't be independently verified.
Auradine's past funding included a $153 million Series C in April 2025, an $81 million Series A in May 2023, and a Series B of more than $80 million in April 2024. The latest round comes from Seligman Ventures, a $500 million fund that launched in February within Seligman Investments.
What Comes Next
Velaura plans to funnel the new capital into accelerating product development, expanding its engineering and customer-facing teams, and deepening relationships with chip vendors and cloud providers. The company unveiled Titan Core in March, positioning the platform as a way to make ultra-low-power computing viable for edge AI while improving data center efficiency.
"Velaura AI is making ultra-low-power computing practical for Physical AI and dramatically more efficient in the data center," said Navin Chaddha, managing partner at Mayfield.
The real test will be whether customers actually adopt the technology at scale and whether the power savings materialize in production. For now, Velaura has the capital and the pedigree. What it still needs to prove is traction.
