Runware Launches Sonic Inference Pods to Scale AI Compute
Runware launched Sonic Inference Pods, modular GPU-heavy data centers designed to reduce AI inference costs and bypass traditional construction delays.
AI inference startup Runware has launched Sonic Inference Pods, modular data centers housed in 20-foot shipping containers. Each pod contains approximately 1,200 GPUs and one megawatt of compute, utilizing closed-loop liquid cooling to enable installation within a single day. The company claims this architecture bypasses the power constraints and construction delays associated with traditional data centers, reducing inference costs by 30 to 90 percent.
Founded by Flaviu Radulescu and Ioana Hreninciuc, the company is currently live in Europe and is deploying assets in the U.S. West. Runware aims to expand to 160 sites, bringing 10,000 nodes online by 2026 and targeting more than one gigawatt of compute capacity by 2027.
Based in London and San Francisco, the startup recently raised a $50 million Series A round led by Dawn Capital, bringing its total funding to $66 million.
Keep reading in the app
The full story and every source, free in the app.
Coming soonGoogle Play