Samsung's IT and cloud unit is offering access to FuriosaAI's AI inference chips.

Samsung SDS has launched an NPU-as-a-Service (NPUaaS) offering in South Korea with FuriosaAI's RNGD NPUs.

The RNGD is now available via the Samsung Cloud Platform, with the companies claiming it to be the first NPUaaS in the country.

FuriosaAI's RNGD accelerators are built on a 5nm Tensor Contraction Processor (TCP) architecture and offer 512 teraflops of FP8 compute. Each accelerator has an 180W thermal design profile.

Each server can have up to eight RNGD accelerators, giving it a 3kW thermal design, which FuriosaAI says enables high inference capacity per rack for agentic AI workloads.

“As organizations deploy increasingly complex AI workloads, they need inference infrastructure that delivers exceptional performance while improving efficiency and lowering the total cost of ownership,” said June Paik, CEO of FuriosaAI. “Making RNGD available through Samsung SDS gives enterprises a new way to deploy AI inference infrastructure without managing dedicated hardware. We’re excited to support Samsung SDS as it expands access to efficient AI inference infrastructure.”

Founded in 2017 and headquartered in South Korea's capital, Seoul, FuriosaAI launched its NXT RNGD server in September 2025.

Samsung SDS is deploying the RNGD accelerators at its data center in Dongtan, and has already virtualized FuriosaAI's NXT RNGD server, offering access to instances in on-demand configurations of 1, 2, 4, or 8 cards.

Samsung SDS, the IT services unit of Samsung, opened its data center in Dongtan, Gyeonggi Province, in March 2023.

Named Dongtan IDC, the data center spans six floors with four server rooms, and has a PUE of 1.12 enabled by its use of natural air cooling. The data center uses the outside wind and colder climate instead of air-conditioning and is hoping to save 21,443 tons of greenhouse gas emissions by 2035.