Runware launches portable modular data center Sonic Inference Pod for edge inference

On Tuesday, AI infrastructure company Runware announced the launch of its own modular data center called Sonic Inference Pod. Designed as a single transportable unit, the Pod represents a more flexible kind of compute that can sit alongside hyperscalers' massive data center projects.

Runware says the Pod can deliver inference at higher quality and lower cost than other serverless inference platforms and GPU clouds. Its modular design makes it easy to add capacity quickly by standing up new pods rather than expanding a fixed data center.

"We believe distributed compute, positioned closer to end users for faster inference, is what will win in the long term," co-founder and CEO Flaviu Radulescu told TechCrunch. Beyond price, the pods can scale fast, deploy anywhere there is power, adapt quickly to new hardware, and use a closed-loop, water-free cooling system that can be built in days, versus the months or even years traditional data centers require.

"Demand for inference is growing faster than facilities can be built," Radulescu said. "What we want is to power the world's intelligence, to be the backbone every AI model runs on with capacity that keeps up with demand instead of throttling it."

Runware currently has 10 pods deployed across the U.S., Europe, and Asia-Pacific, and already provides inference to companies including Higgsfield AI and Wix, with 160 sites available to power its pods. The company raised a $50 million Series A in December, led by Dawn Capital and Comcast Ventures, to fund infrastructure for image generation; it sees the expansion into pods as part of its core mission of providing inference to companies rather than a single product.

Meanwhile, AI labs like OpenAI and SpaceX are still racing to build data centers across the U.S. — OpenAI is reportedly close to a $500 billion deal to build a data center in Ohio. Radulescu doesn't see those projects as a threat, describing the pods' flexibility as the key differentiator.

"Every pod runs as part of a single network, so requests go wherever there's capacity, closer to the users, and if one pod goes offline, traffic moves to another," he said, adding that a system failure means one pod is down rather than a whole fixed facility. Customers who want dedicated hardware get whole pods to themselves.

The watch point: AI data centers are controversial largely because of the resources they consume. Whether Runware's small, fast-to-deploy pods can thread the needle on cost, efficiency, and public scrutiny will determine whether its promise of keeping capacity ahead of inference demand holds up.