Texas-based modular data center company Velox DCI Europe has partnered with AI infrastructure firm I/ONX HPC for the deployment of more than 7,000 FuriosaAI RNGD accelerators at a data center in Stockholm, Sweden.

Velox broke ground on the facility earlier this year, with the first 2MW of compute capacity expected to come online in early 2027. The company is aiming to add a further 8MW of capacity later that same year, targeting an overall total capacity of 15MW.

The RNGD accelerators will be deployed in multiple phases, with an initial 1,800 installed in Phase I, with a total of 7,000 accelerators deployed across “subsequent expansion phases.”

In a statement, Furiosa, a South Korean AI chip startup, said its hardware will operate alongside GPUs deployed as part of I/ONX high-density Symphony SixtyFour platform. According to I/ONX’s website, the Symphony SixtyFour platform “consolidates up to 64 accelerators into a single node, reducing rack-scale TCO by 50 percent or more.”

In a post on LinkedIn, I/ONX CEO Steven Eliuk said the 7,000 RNGD NPUs would sit alongside approximately 7,000 GPUs from Nvidia and AMD.

Velox will develop, manage, and operate the physical data center facility, which will then serve customers under an Infrastructure-as-a-Service model. I/ONX, meanwhile, will serve as the system integrator and platform partner, with Furiosa supplying the RNGD accelerators, drivers, software stack, and architectural support.

“Over the coming months, the teams will continue technical validation, software integration, and supply planning in preparation for the first deployment phase,” Furiosa said in a statement. “We are honored to work alongside I/ONX and Velox to help build the foundation for sustainable, high-performance AI compute in Europe.“

Founded in 2017 and headquartered in South Korea’s capital, Seoul, to date, FuriosaAI has raised approximately $115 million across four funding rounds to support the development of its RNGD chip. Furiosa’s 180W, PCIe-based RNGD accelerators are currently in mass production, fabricated on TSMC’s 5nm process node.

In March 2024, the startup turned down an $800m acquisition offer from Meta, with the startup reportedly wanting to grow its business as an independent company instead. In January 2026, it was reported that the company was seeking to raise between $300 and $500 million in a funding round ahead of its anticipated IPO.

In May 2026, the chip startup announced it was partnering with Broadcom for the development of its third-generation AI accelerator, designed to support inference workloads.

Velox, meanwhile, launched in February 2025, with the company saying it enables “the world’s leading data center builders to scale capacity faster — with greater certainty on schedule, cost, and quality.” On its website, Velox claims it can achieve “up to 1GW deployments in as little as 51 weeks under the right conditions.” The project in Sweden appears to be its first deployment.

Founded in 2020 and headquartered in Las Vegas, Nevada, I/ONX describes itself as a “leading provider of heterogeneous AI infrastructure.” In addition to the Symphony SixtyFour, the company also offers Synth, which puts “data center-grade heterogeneous compute in a suitcase-sized package,” and the Canon storage rack.