Velaura AI, an AI compute infrastructure company, has raised $110 million in a Series A funding roun...

Velaura AI, an AI compute infrastructure company, has raised $110 million in a Series A funding round led by Seligman Ventures, with participation from existing investors including Amplify Partners and Wing Venture Capital. The round pushes Velaura's valuation past the $1 billion mark, a significant milestone for a company focused on energy-efficient hardware for AI inference workloads.

The company specializes in FPGA-based accelerators designed to reduce the power consumption of large-scale AI inference. Its flagship Velos platform targets enterprise customers running models in production, offering configurable hardware that can be tuned for specific model architectures and batch sizes. Velaura claims its systems deliver up to five times better energy efficiency compared to general-purpose GPUs for certain natural language processing tasks, a claim that positions the company squarely in the growing 'sustainable AI' segment of the infrastructure market.

This funding round arrives at a time when data center operators are grappling with surging electricity demand from AI workloads, and enterprises are increasingly prioritizing power efficiency in their hardware procurement decisions. Velaura's approach aims to address this by providing an alternative to power-hungry GPUs for inference-heavy deployments, particularly in scenarios where latency and throughput requirements can be met by specialized silicon.

The company states that the new capital will be directed toward scaling its hardware roadmap, expanding its software toolchain, and growing go-to-market teams in North America and Europe. Velaura also plans to invest in its large language model optimization toolkit. While the company did not disclose a specific customer list in the announcement, it indicated that its platforms are already in use across several enterprise environments, including in healthcare and financial services, where energy constraints and data sovereignty requirements make on-premise inference attractive.

The investment signals increasing investor appetite for compute alternatives that address operational and sustainability challenges rather than raw performance alone. As model deployments shift toward inference at scale, efficiency-driven hardware sits at the center of the next phase of AI deployment economics.