Thunder Compute, a San Francisco-based startup, has secured $13 million in Series A funding to advance its mission of virtualizing GPUs and putting idle compute to work. The round was led by Matrix Partners, with participation from Y Combinator and CEAS Investments. The company estimates that $200 billion worth of compute is wasted annually due to low GPU utilization rates, which average only about five percent in data centers.
The funding will enable Thunder Compute to scale its proprietary GPU virtualization software, which treats GPUs as network resources and operates invisibly beneath workloads to boost data center efficiency. By virtualizing GPUs, the company aims to unlock idle capacity and provide additional compute without the need for new hardware purchases. This approach could significantly reduce both costs and environmental impact for enterprises.
Thunder Compute's technology is designed to be transparent to applications, allowing data centers to pool and share GPU resources dynamically. This can increase utilization rates and reduce the need for over-provisioning, which is a common practice to meet peak demand. The company plans to partner with enterprises to deploy its solutions at scale, targeting a future where every GPU is virtualized.
The startup was founded in 2022 by Carl Peterson, previously a management consultant at Bain & Company, and Brian Model, a former quantitative developer at Citadel Securities. Their backgrounds in business strategy and quantitative finance bring a unique perspective to the infrastructure challenges facing the AI industry. The company is also backed by Y Combinator, having gone through its accelerator program.
The announcement comes amid growing concerns about GPU shortages, particularly for AI workloads. By improving utilization of existing hardware, Thunder Compute offers a complementary approach to building new data centers. This could be especially impactful for small and medium-sized enterprises that struggle to access high-performance computing due to cost and availability constraints.
According to the company, the funding will accelerate product development and expand its team. They have already demonstrated significant interest from potential customers. The investment from Matrix Partners and CEAS Investments underscores the market's confidence in the potential of GPU virtualization to address a critical bottleneck in the tech industry.
For more information, visit the company's website at Thunder Compute.
