Velaura Raises $110M After Pivoting From Crypto Chips to AI

Reuters’ account of the financing says Velaura AI raised $110 million in a Series A on August 18, 2026, at a valuation above $1 billion, with Seligman Ventures leading the round and Capricorn Investment Group joining as a new investor.

The financing places a unicorn valuation on a recent change in direction. Formerly known as Auradine and focused on bitcoin-mining chips, the Santa Clara semiconductor company is now seeking to commercialize low-power design technology for AI accelerators through its Titan Core offering.

The round finances development, not proven AI sales

Investors are funding the expansion of an AI-compute portfolio whose commercial record is still emerging. The capital is intended for product development, commercialization, additional engineering and customer-facing staff, and closer work with partners and prospective customers.

In Velaura’s official financing announcement, the company claims Titan Core improves performance per watt by two to four times for mathematical operations used in AI accelerators, states that its underlying technology has been deployed in more than 30 million production ASICs, and identifies multiple hyperscaler engagements.

The announcement does not provide the precise valuation, investor ownership, liquidation preferences, board changes or a schedule for deploying the proceeds. Nor does it allocate the capital between data-center products and technology for robots, drones and other systems operating under tight power and thermal limits.

Those gaps define what the financing establishes. It demonstrates investor willingness to back Velaura’s technology and strategy, but it is not evidence of revenue from Titan Core, completed customer qualification or volume shipments of an AI product.

Auradine’s mining-chip history underpins the pivot

A SiliconANGLE account of the transition traces Velaura to the former Auradine bitcoin-mining chip business, dates the move toward Titan Core to March 2026—about six months before the financing—and describes the new offering as processor building blocks and engineering services for customers developing AI chips.

Titan Core is therefore better understood as a silicon design and intellectual-property platform than as a complete general-purpose AI processor. Its components include low-voltage cell libraries, custom circuit development and a proprietary toolflow intended to support chip reliability and manufacturing yield.

The mining business gives Velaura relevant experience in power management, circuit implementation, thermal constraints and ASIC production. That history matters because semiconductor technology must survive fabrication and operate reliably outside a laboratory, but bitcoin-mining deployments do not automatically validate demand from AI-accelerator customers with different workloads, software stacks and qualification processes.

Production deployment is not the same as AI adoption

The strongest evidence behind Velaura’s pitch concerns the underlying design technology, not the market performance of Titan Core. Prior deployment in production ASICs indicates that elements of the low-power foundation have encountered real manufacturing, yield and reliability requirements at scale.

It does not establish that Titan Core-based AI chips have shipped at comparable volume. The production figure is attached to underlying technology and should not be read as a shipment count for the newly targeted AI platform.

The efficiency claim also has a narrower scope than a system-level comparison. It concerns specified mathematical operations inside an accelerator; total energy use also depends on memory, networking, software behavior, cooling and the way a complete chip is integrated into a server or physical-AI device.

No independently reproduced benchmark, detailed test configuration or named production AI customer accompanied the financing. Until such evidence appears, the performance claim remains a company-supplied result rather than an independently verified measure of a complete accelerator or deployed system.

Hyperscaler engagement remains commercially undefined

Work with hyperscalers could become important validation, but the public description does not identify the companies or define the stage of those relationships. An engagement can range from an early technical evaluation to paid design work, a pilot, a committed production program or a completed deployment.

Investors are consequently backing two propositions with different levels of support. Velaura has a manufacturable low-power technology base with prior ASIC deployment, while the proposition that customers will integrate Titan Core into commercially successful AI products at scale remains unproven in the public record.

Power efficiency gives the company a relevant competitive target because data centers, robots and autonomous systems operate within electrical and thermal limits. Commercial success will depend on whether the claimed advantage survives integration into a complete product without unacceptable costs in silicon area, software compatibility, memory performance or development time.

The next evidence must come from AI products

The milestones that would narrow the gap between technical history and AI commercialization are named design wins, completed tape-outs and independently documented performance-per-watt results on representative workloads. Production qualification, shipment schedules and revenue specifically attributable to Titan Core would provide stronger evidence than an undisclosed customer engagement.

For now, the financing validates investor appetite for moving Velaura’s production-experienced semiconductor team into AI computing. The underlying low-power design foundation has a deployment history; public proof that Titan Core itself is a shipping, broadly adopted AI offering has yet to follow.

Also read:

• The Death of a VR Unicorn: Rec Room Shuts Down After 150 Million Users, $294 Million Raised, and a $3.5 Billion Valuation
• Crypto's Hidden Liquidity Trap: Tom Lee's Bombshell on Market Makers and the Bitcoin Bloodbath
• The Crypto Market Is Dead: No Retail Buyers Left, Faith Is Gone, and 6,000 Out of 8,000 Projects Are Pure Scams