Velaura AI Secures $110M to Disrupt Data Center Power Economics

Avatar photo

ByLisa Grant

August 19, 2026

Velaura AI raised $110 million to scale its ultra-low-power Titan Core platform, targeting the massive energy costs currently burdening major cloud providers and AI developers.

The escalating energy demands of the algorithmic state have met a new challenger. Velaura AI, a semiconductor startup focused on ultra-low-power silicon, announced a $110 million Series A funding round on August 18, 2026. Led by Seligman Ventures and supported by the Samsung Catalyst Fund, Capricorn Investment Group, and Maverick Silicon, the round values the company at over $1 billion. This capital injection targets the primary bottleneck of modern surveillance and data capitalism: the unsustainable power consumption of AI data centers that currently drives up costs for every developer using AWS, Google Cloud, or Linode.

Velaura’s proprietary Titan Core platform is designed for “physical AI” workloads, including robotics and autonomous systems. The company’s business model reflects a shift toward performance-based accountability, utilizing a royalty structure tied directly to a share of the customer’s power savings. This approach explicitly aligns the startup’s profits with the user’s ability to reduce overhead. With engagements already underway with three of the four largest cloud providers, Velaura’s technology is already deployed in more than 30 million chips, signaling a behind-the-scenes shift that could lower inference costs for those utilizing Anthropic, OpenAI, or ElevenLabs via major cloud marketplaces.

As hardware efficiency evolves, software giants are moving to consolidate control over AI spending. Snowflake announced the launch of its Cortex AI Gateway, featuring dynamic model routing derived from its Natoma acquisition. This technology allows enterprises to define cost and quality parameters, automatically selecting the most economical model for a given task using a feedback loop. By integrating models like DeepSeek-V4-Flash and GLM-5.3 into its private preview catalog, Snowflake is positioning itself as a centralized governance layer for the fragmented AI market, providing a native alternative to third-party tools like OpenRouter.

This trend toward centralized AI management is further evidenced by F5’s new Agentic-Ready AI Gateway. Designed to track “AI tokenomics,” the platform provides a unified control point for monitoring usage and spend across various model providers. For organizations managing multiple API keys and GitHub-hosted services, these gateways represent a double-edged sword: they offer necessary cost controls and security against AI-discovered vulnerabilities while further centralizing data flow through major infrastructure vendors. This comes as Palo Alto Networks also enters the fray with its Frontier AI Critical Defense Program, aiming to protect critical infrastructure against increasingly sophisticated digital threats.

While hardware and software providers race to optimize, the broader economic landscape remains volatile. U.S. government interest costs have reached an annualized $1.2 trillion, surpassing defense spending, while bond yields recently touched 6% before Treasury intervention provided a temporary reprieve. Despite these fiscal pressures, capital continues to flow into the AI sector. Natural recently secured a $100 million credit facility from Upper90 Capital Management to scale payments infrastructure specifically for AI agents, signaling that the transition toward an automated economy is accelerating despite growing worker uncertainty regarding job displacement.

The market for these protective and efficient technologies is expanding rapidly. The critical infrastructure protection market is projected to reach $206.31 billion by 2031, while network-attached storage is expected to surge to nearly $97.44 billion by 2032. Even as OpenAI pauses some model work over safety concerns, the momentum behind Anthropic and open-source alternatives suggests that the push for digital sovereignty will depend on whether these new efficiency-focused technologies can truly democratize access to compute or if they will simply further entrench the existing cloud giants.

Leave a Reply

Your email address will not be published. Required fields are marked *