On Tuesday, AI infrastructure company Runware unveiled its modular data center, the Sonic Inference Pod. Designed as a single transportable unit, the pod offers a more flexible approach to compute, positioned to complement—not replace—the massive data center projects of hyperscalers.
According to Runware, the pod delivers inference at higher quality and lower cost compared to other serverless inference platforms and GPU clouds. Its modular architecture allows for rapid capacity expansion by simply adding new pods, avoiding the need to expand fixed facilities. In many ways, this is the future, said Flaviu Radulescu, co-founder and CEO of Runware, in an interview with TechCrunch.
“We believe distributed compute, positioned closer to end users for faster inference, is what will win in the long term,” Radulescu said, pointing to his company as an example. Beyond cost savings, he highlighted that the Runware system can scale and add capacity quickly, deploy anywhere with power access, and adapt rapidly to new hardware releases. The pods also use a closed-loop cooling system that requires no water, and can be built in days—compared to the months or even years needed for traditional data centers.
“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. The company already provides inference services to clients including Higgsfield AI and Wix, and has 160 sites available to power its pods. In December, Runware announced a $50 million Series A round to build the infrastructure needed for companies to generate images. The expansion into pods aligns with the company’s core mission: providing inference as a service, rather than a single product.
As AI labs like OpenAI and SpaceX continue racing to construct data centers across the U.S., the Sonic Inference Pod represents a complementary—and potentially disruptive—alternative. For instance, OpenAI is reportedly nearing a $500 billion deal to build out its own infrastructure. In this context, modular and portable data centers could offer a pragmatic path to meeting the surging demand for AI inference, enabling deployment where it’s needed most.
via TechCrunch AI