Thunder Compute, a startup founded in 2022, has secured $13 million in Series A funding to address a critical inefficiency in the data center industry: the vast amount of idle GPU capacity. The round was led by Matrix Partners, with participation from Y Combinator and CEAS Investments, according to an announcement made on August 19, 2026.
The company’s proprietary GPU virtualization software treats GPUs as network resources, operating invisibly beneath workloads to boost data center efficiency. This approach aims to tackle a significant problem: average GPU utilization is only about five percent, meaning that nearly $200 billion of compute capacity is wasted globally. By virtualizing GPUs, Thunder Compute seeks to unlock this idle capacity and provide additional compute power without requiring new hardware.
The funding will be used to scale the company’s operations and partner with enterprises to virtualize GPUs at scale. Thunder Compute’s technology could have major implications for the AI and machine learning sectors, where GPU demand often outpaces supply, leading to long wait times and high costs. By efficiently utilizing existing GPUs, the company could help alleviate the capacity shortage and reduce the environmental impact of manufacturing new hardware.
Thunder Compute was founded by Carl Peterson, formerly a management consultant at Bain & Company, and Brian Model, previously a quantitative developer at Citadel Securities. Their background in business strategy and quantitative finance brings a unique perspective to the technical challenges of GPU virtualization.
The company’s mission is to create a future where every GPU is virtualized, enabling data centers to operate at much higher efficiency levels. This could lead to significant cost savings for companies that rely on GPU-intensive workloads, as well as a reduction in the carbon footprint of the tech industry. The announcement highlights the growing importance of resource optimization in the face of increasing demand for computational power.
With the new funding, Thunder Compute plans to expand its team and accelerate product development. The company’s software is designed to be transparent to existing applications, making it easier for data centers to adopt without major disruptions. This could lower the barrier to entry for organizations looking to improve their GPU utilization rates.
The investment from prominent firms like Matrix Partners and Y Combinator signals confidence in Thunder Compute’s approach. As the demand for GPUs continues to rise, solutions that maximize the use of existing hardware will become increasingly valuable. Thunder Compute’s success could pave the way for more sustainable and cost-effective computing practices across the industry.
For more information, visit Thunder Compute’s website at thundercompute.com.
