❯ FuriosaAI closes $620 million Series D for inference-chip production and next-generation development

Capital follows the start of volume productionSouth Korean AI-chip developer FuriosaAI announced a $620 million Series D at a $2.9 billion post-money valuation this week. Its release names investors including Cisco Investments, Infinitum and Jericho Capital. Korea’s National Growth Fund committed KRW370 billion, about $280 million, as part of the round rather than an amount to add on top.

Hardware and software built for inferenceFounded in 2017, FuriosaAI develops inference accelerators and compiler software. It supplies cloud providers and enterprises with hardware and the accompanying capabilities to run trained models in production. Its second-generation RNGD comes as a 180-watt, air-cooled PCIe card, targeting customers with existing racks, power and cooling and reducing the need for infrastructure retrofits.

Deployments provide a commercial foundationThe company says RNGD entered volume production in January 2026. Samsung SDS has launched an NPU service using it, while LG Group companies are deploying it in on-premises and hybrid-cloud environments. Furiosa plans to produce 20,000 cards this year and says it has secured memory and foundry capacity to more than double production in 2027. The company is arranging its supply chain around that target; full-year production and shipments remain to be reported.

Funding the next stage of deliveryThe Series D will expand global RNGD production and deployments and fund a third-generation chip for large clusters running frontier agentic models. SDK, caching, batching and Kubernetes improvements aim to make existing workloads easier to move. In competition with NVIDIA, card performance is one requirement among several: enterprises also need model compatibility, maintainable software and reliable supply. The financing supports that wider delivery capability.

▮ SIGNALCommercial proof for inference chips now spans production, software compatibility and customer deployment, requiring capital across the delivery chain.