FuriosaAI, a South Korean artificial intelligence chip company, has established a local subsidiary in Singapore. The strategy is to use Singapore as its hub for the Asia-Pacific region and accelerate global commercialization of RNGD, the company's second-generation AI inference accelerator.
FuriosaAI said on the 11th that it has set up FuriosaAI Singapore Pte. Ltd. and will run it as a core base handling business development, sales and technical support across the Asia-Pacific region. The Singapore unit will be led by Addison Chee, vice president of APAC sales at FuriosaAI. With the new subsidiary, the company's global footprint now spans Silicon Valley in the United States, Lisbon in Portugal and Singapore, in addition to its Seoul headquarters.
Building on the local unit, the company plans to work with businesses, governments, cloud providers and infrastructure partners in Southeast Asia and elsewhere in the Asia-Pacific region to drive large-scale adoption of AI inference infrastructure. It also plans to keep adding local staff and strengthen its technical support capabilities so it can work closely with customers from initial technology validation through to building actual commercial environments.
FuriosaAI is expanding global commercial deployments of RNGD, an AI accelerator for data centers. RNGD is a second-generation NPU designed to run a range of AI workloads, including large language models, multimodal models and agentic AI, with high performance and efficiency. Mass production began in January, and the chip has been adopted by major domestic and overseas customers, including Samsung SDS and LG AI Research, for use in commercial services.
"The Asia-Pacific region is rapidly growing into a core market for global AI infrastructure as companies deploy advanced AI within the real power and infrastructure constraints of their data centers," said Baek June-ho, chief executive of FuriosaAI. "Through our Singapore base, we will work more closely with customers and partners in the region and support them so that RNGD moves smoothly from technology validation to actual commercial deployment."
