FuriosaAI secures $620 million for AI chip production and its next processor
The financing backs a chip already in mass production, a larger global rollout and a third-generation design under development with Broadcom.
Loading page…
The financing backs a chip already in mass production, a larger global rollout and a third-generation design under development with Broadcom.
Listen to this story
FuriosaAI’s $620 million Series D, announced October 28, 2026, values the company at $2.9 billion and gives it capital to scale an inference-chip business that has moved into mass production. The financing is anchored by South Korea’s National Growth Fund, which invested about $280 million, while the company’s commercial footing includes Samsung SDS and LG deployments. The key test now is whether Furiosa can deliver its 2026 production target and translate those deployments into a larger business as it develops a more advanced chip with Broadcom and other partners.
FuriosaAI targets 20,000 RNGD cards in 2026 and says secured memory and fabrication capacity could more than double output in 2027.
Samsung SDS offers access to RNGD computing, while LG Group companies are deploying the cards in on-premises hybrid-cloud environments.
The planned third-generation accelerator includes a 2-nanometer, multi-die design and HBM4/4E memory; Furiosa is developing it with Broadcom and other partners.
FuriosaAI now has $620 million to turn its production foothold into a larger AI chip business while developing its next processor. The Seoul-based company announced the Series D on October 28, 2026, at a $2.9 billion post-money valuation. It says the capital will fund more chip production, global deployments, software development and a third-generation accelerator.
Furiosa says its second-generation RNGD accelerator entered mass production in January 2026. The chip is built specifically for inference—the work of running a trained AI model to produce an answer. The company plans to produce 20,000 cards in 2026 and says it has secured memory and chip-fabrication capacity to more than double output in 2027.
Samsung SDS has launched a service offering access to RNGD computing, while LG Group companies are deploying the hardware in on-premises hybrid cloud environments, according to Furiosa.
RNGD comes as a 180-watt, air-cooled card. Eight cards make up Furiosa’s NXT RNGD Server, a 3-kilowatt system designed to fit existing enterprise racks without a cooling retrofit. Furiosa claims RNGD matches RTX PRO 6000 throughput at the same card count while serving twice as many users per kilowatt.
We look for the moment a technology company moves from proving its technology to proving its business. FuriosaAI is at that point now, with its chip in volume production and in deployment with customers.
Yong Jung Park, General Partner, NAVER Ventures
South Korea’s National Growth Fund was the round’s largest investor, putting in about $280 million, according to the company’s financing release. The release links that investment to the country’s efforts to expand AI computing infrastructure and strengthen its semiconductor ecosystem.
International investors supplied $110 million, Bloomberg reported. Participants included Infinitum, Cisco Investments, Jericho Capital, QUDA Capital, NeoFort Holding and Worldline Capital. All three original seed investors—NAVER, DSC Investment and KDB Capital—also returned, with NAVER participating through NAVER Ventures.
Beyond the current production ramp, Furiosa is developing its third-generation accelerator with Broadcom and other partners. It says the next chip will retain its Tensor Contraction Processor architecture. The planned design includes:
Software is another destination for the proceeds. Furiosa says it has shipped four major software development kit releases this year. It also says customers can deploy through PyTorch as they would on GPUs, while developers seeking finer control can write their own low-level routines.
Loading discussion...
Join the conversation
Explain where you would draw the line between national investment and private risk.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.