top of page

Thunder Compute Plans to Put Idle GPUs Back to Work

  • Writer: Karan Bhatia
    Karan Bhatia
  • 5 hours ago
  • 2 min read

Thunder Compute, building a new way to virtualize GPUs, led by Carl Peterson and Brian Model, has raised a $13 million Series A round led by Matrix Partners, with participation from Y Combinator and CEAS Investments. The capital will be deployed to address the GPU capacity shortage by eliminating the estimated $200 billion in wasted compute currently sitting idle, while building toward a future in which every GPU is virtualized.


Enterprises currently underutilize GPUs, with average utilization at just five percent, according to Cast AI. Unlike CPUs, storage, and memory, GPUs typically remain allocated to a single workload even when idle. Thunder Compute aims to change this by virtualizing GPUs, allowing multiple workloads to share the same physical capacity.


Thunder Compute co-founder Carl Peterson said more than $200 billion in data center capacity sits idle because GPUs lack the virtualization available for other hardware. The company has developed GPU virtualization technology and tested it through a self-serve cloud used by more than 10,000 users. The new funding will support scaling the technology for enterprise GPU fleets.


Thunder Compute’s proprietary software treats GPUs as network resources, allowing workloads to access them whenever and wherever needed. By operating beneath the workload layer, the technology virtualizes GPUs without requiring changes to existing code or workflows. This allows enterprises to improve GPU utilization while keeping the experience seamless for developers.


Matrix Partners General Partner Ilya Sukhar said Thunder Compute stands out by taking a generalized approach to data center optimization rather than targeting individual workloads. He highlighted the team’s foresight and said the timing is right to deploy its technology to address the growing GPU capacity crunch.


𝐑𝐞𝐦𝐨𝐭𝐞 𝐇𝐢𝐫𝐞 𝐖𝐢𝐭𝐡 𝐔𝐬: 𝐌𝐞𝐧𝐥𝐨 𝐓𝐚𝐥𝐞𝐧𝐭 helps technology companies hire exceptional remote talent from India. Build your team with carefully curated engineers, product leaders, designers, GTM professionals, and more. Learn More At: https://www.menlotimes.com/menlo-talent

Menlo Times is a global media platform covering AI, Deeptech, Venture Capital, Fintech, Robotics, and Security through news, analysis, and insights from founders and operators.
  • Instagram
  • Facebook
  • X(Formerly Twitter)
  • LinkedIn
  • YouTube
© 2026 Menlo Times. All rights reserved.
bottom of page