Velaura AI secured $110 million in Series A funding at a valuation exceeding $1 billion to advance its Titan Core ultra low power silicon platform for AI data centers and Physical AI systems.
Velaura AI (formerly Auradine) closed a $110 million Series A funding round, valuing the Santa Clara-based company at more than $1 billion (with some reports citing approximately $1.05 billion). The round was led by Seligman Ventures, with new participation from Capricorn Investment Group and Prosperity7 Ventures, alongside existing investors including Mayfield, Maverick Silicon, MARA, Premji Invest, Samsung Catalyst Fund, and StepStone Group.
Proceeds will accelerate development and commercialization of the company’s AI compute portfolio (centered on the Titan Core platform), expand engineering and customer facing teams, and deepen engagements with strategic partners and customers in next generation AI infrastructure and Physical AI.
What is Velaura AI?
Velaura AI originated as Auradine, founded in 2022. It initially focused on energy efficient silicon and systems for blockchain (notably Bitcoin mining ASICs and related infrastructure) and early AI applications, raising substantial prior capital, including an oversubscribed $153 million Series C in April 2025 (largely equity with some debt) that brought total funding to over $300 million. Auradine generated meaningful revenue from mining hardware (annualized run rate exceeding $150 million reported around the Series C) and had deployments with dozens of operators.
In March 2026, the company underwent a major restructuring and rebranded as Velaura AI. Core engineering and silicon technology remained intact, but the strategic emphasis shifted fully toward ultra low power compute for cloud/hyperscale data centers, edge, and Physical AI (robots, drones, autonomous systems, and other embodied intelligence). New leadership was added, including Manu Gulati as co-founder and chief development officer, with Aditya Grover as founding advisor. The leadership and engineering team draws heavily from Apple, NVIDIA, Google, Qualcomm, Marvell, and other semiconductor veterans; co-founder and CEO Rajiv Khemani is a serial entrepreneur previously associated with Innovium (acquired by Marvell) and Cavium.
This pivot positions Velaura to address the binding constraint of power availability and cost in AI scaling, rather than competing primarily on raw performance or clock speed. The underlying low power technology has already been validated at commercial scale, deployed across more than 30 million ASICs on leading process nodes with strong yield and reliability.
Core Technology: Titan Core
Titan Core is Velaura’s proprietary digital chip IP and silicon design platform, unveiled in March 2026. It targets the arithmetic/matrix multiplication operations that typically account for ~40–70% of power in an XPU/GPU. Using patented ultra low power techniques (including low voltage libraries, proprietary circuit design, and specialized EDA flows), it claims a 2–4× reduction in power for those operations while maintaining full functional performance and equivalence.
Key claimed outcomes for a typical 1,000 W XPU/GPU include:
- Overall chip power reduction of up to ~2× (up to 250–500 W savings).
- Electricity cost savings of roughly $650–$1,300 per chip over three years (at illustrative rates; excluding cooling and infrastructure).
- Potential tens to hundreds of millions of dollars in annual savings at data center scale, plus lower cooling needs, higher rack density within existing power envelopes, and the ability to deploy more compute where power is constrained.
The engagement model is IP/design services oriented (not a merchant silicon product). Customers provide RTL and priorities; Velaura applies its IP, libraries, tool flows, and expertise to deliver optimized GDS or chiplets. It integrates into existing SoC architectures and design flows. The business model mirrors aspects of Arm’s historical licensing approach: an upfront fee plus a royalty structured as a share of the power savings achieved. Velaura is engaged with multiple hyperscalers (reported as three of the four largest cloud providers) and XPU/GPU vendors on 3 nm and 2 nm nodes, though specific identities and commercial deployment status remain undisclosed.
The technology is also being extended to Physical AI use cases, where tighter energy and thermal budgets make efficiency gains particularly valuable for battery life, always-on operation, and autonomy.
AI infrastructure buildout is increasingly power constrained. Hyperscalers are committing hundreds of billions of dollars to data centers, yet face multi year delays for power availability, grid connections, and cooling. Electricity and total cost of ownership (TCO) have become first order competitive factors alongside model quality and raw FLOPS. Velaura’s thesis is that the next era of AI will be defined by superior compute economics, more useful work per watt, rather than solely by larger models or higher peak performance.
Recommended: Social Media Leader ByteDance Joins the Open Invention Network Community 2.0
This creates dual opportunities:
- Hyperscale data centers: higher effective capacity within fixed power budgets, lower operating costs, and accelerated deployment.
- Physical/edge AI: enabling denser, longer running intelligent systems in robots, drones, vehicles, and other embodied platforms.
Investor interest reflects conviction in the technology’s production track record, early hyperscaler traction, the experienced team, and the structural power bottleneck.
Capital will primarily fund:
- Acceleration of Titan Core commercialization and productization.
- Expansion of engineering capacity and customer success/sales teams.
- Support for ongoing and new hyperscaler and Physical AI collaborations.
The company aims to convert design engagements into production deployments. Success metrics will center on measurable, workload specific power savings, seamless integration without software disruption, and repeatable TCO improvements that customers can underwrite at scale.
Velaura operates in a crowded field of AI accelerator and efficiency efforts (NVIDIA, AMD, custom ASICs from hyperscalers, other IP and chiplet players). Its differentiation rests on production proven low power digital techniques applied specifically to arithmetic intensity, a licensing model that shares value with customers via savings, and dual focus on cloud and Physical AI. Risks include the need to demonstrate consistent real world savings beyond lab or prior ASIC results, competition from process node advances and architectural innovations by incumbents, customer qualification timelines in high stakes data center environments, and execution on scaling the IP/services business while potentially managing any residual legacy operations from the Auradine era.
The $110 million round and unicorn valuation mark a clear capitalization of Velaura’s pivot toward solving AI’s power economics at the silicon level, backed by a technology foundation already shipping in volume and early commercial interest from major cloud providers.
Please email us your feedback and news tips at hello(at)techcompanynews.com
