FuriosaAI Secures $620 Million Series D to Meet Surging Enterprise AI Inference Demand and Accelerate Next-Gen Silicon

FuriosaAI, an AI semiconductor company building sustainable, high-performance inference compute for data centers, today announced the completion of its $620 million (KRW 820 billion) Series D round, bringing its post-money valuation to approximately $2.9 billion.

The financing marks a major milestone as FuriosaAI scales global commercial deployments of its RNGD AI inference accelerator. FuriosaAI will use the new capital to expand RNGD production, strengthen its software ecosystem, and accelerate development of its third-generation AI accelerator.

International investors contributed a total of $110 million to the round, bringing deep expertise across AI semiconductors and computing infrastructure. U.S.-based Infinitum, an investment firm with extensive experience in the AI semiconductor sector, was the largest international investor in the round.

Singapore-based QUDA Capital Partners also joined the round. Its Founder and Managing Partner, Charlie Foo, previously held senior leadership roles at NVIDIA and Mellanox Technologies, including as Vice President and General Manager for Asia Pacific Enterprise Business. He currently serves as Principal Advisor to FuriosaAI.

Other international investors include Cisco Investments, Abu Dhabi-based NeoFort Holding, and U.S.-based technology investment firms Jericho Capital and Worldline Capital.

South Korea's National Growth Fund invested $280 million as the largest investor in the round, supporting the country’s initiatives to expand AI computing infrastructure and strengthen the Korean AI semiconductor ecosystem, while helping accelerate the round's momentum.

Existing shareholders contributed $125 million in follow-on investments, reflecting continued confidence in FuriosaAI's long-term growth prospects. Notably, the Korea Development Bank (KDB), one of FuriosaAI's major existing shareholders, invested an additional $22 million in the round.

All three of FuriosaAI’s original seed investors, NAVER, DSC Investment, and KDB Capital, participated. DSC Investment, which first backed FuriosaAI in 2017, has participated in every financing round since the company's seed stage. NAVER's seed investment was made by NAVER D2SF, its early-stage investment arm, while this round's investment was made by NAVER Ventures, NAVER's Silicon Valley-based investment platform.

"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," said Yong Jung Park, General Partner of NAVER Ventures. "As a Silicon Valley-based investor, helping the company grow in North America is where we can add the most value. We're proud to back June and the team as they take Furiosa to the global stage."

Nextrans and Kakao Investment also participated. Both firms became FuriosaAI shareholders in its Series C bridge round last year and increased their commitments as the company advanced mass production and global commercialization. The round also attracted new Korean institutional investors. The Export-Import Bank of Korea selected FuriosaAI for its first direct investment in a startup. Korea’s United Asset Management Company (UAMCO) committed $30 million, becoming the largest new domestic investor after the National Growth Fund, while Hanwha Asset Management also joined the round.

“This Series D financing marks an important milestone for FuriosaAI and reflects the confidence of investors around the world in our technology and ability to execute,” said June Paik, Co-Founder and CEO of FuriosaAI. “We are particularly encouraged by the participation of investors with deep expertise and experience in the global AI semiconductor and infrastructure industries.”

“We are entering a new phase of bringing our technology to the global AI infrastructure market at scale,” Paik added. “By accelerating the commercialization of RNGD and advancing the development of Gen 3, we aim to help shape the next generation of AI infrastructure.”

“Taking an advanced AI chip from design to volume production is an exceptionally difficult undertaking, and few companies have demonstrated the ability to do so successfully,” said John Yetimoglu, Chief Investment Officer with Infinitum. “FuriosaAI has gone further, securing production deployments with leading enterprises such as Samsung SDS and LG Group. This combination of technological innovation and commercial execution gives us strong conviction in FuriosaAI's ability to scale RNGD globally and deliver on its ambitious Gen3 roadmap.”

FuriosaAI expects to produce 20,000 RNGD accelerators in 2026, with plans to more than double production to over 40,000 units in 2027, supporting our growing customer demand. LG Group is running its EXAONE deployments on RNGD, Upstage now uses RNGD to serve live search summarization for millions of users, and Samsung SDS launched its commercial NPU-as-a-service on RNGD in July.

About FuriosaAI

Founded in 2017, FuriosaAI is a South Korea–based semiconductor company developing high-performance, energy-efficient AI accelerators for data center inference. Leveraging its proprietary Tensor Contraction Processor (TCP) architecture, FuriosaAI designs and develops an integrated AI computing stack spanning silicon, compilers, software, and server systems, with a mission to make AI computing more sustainable and accessible.

FuriosaAI's second-generation accelerator, RNGD, entered mass production in January 2026. Featuring high-bandwidth memory and a power-efficient architecture, RNGD delivers scalable AI inference within existing data center power and cooling constraints. The company is expanding commercial deployments with enterprise and cloud customers globally, while collaborating with Broadcom on its third-generation AI accelerator for hyperscale infrastructure.

Headquartered in Seoul, FuriosaAI has offices in the United States, Portugal, and Singapore.

Learn more at furiosa.ai.

About Furiosa’s Tensor Contraction Processor: A new chip architecture designed for performance, efficiency, and programmability

GPUs and most other AI accelerators are designed to break model operations into fixed-size, two-dimensional matrix multiplications. Furiosa’s Tensor Contraction Processor (TCP) instead makes a higher-dimensional operation, tensor contraction, the basic unit of work for both the hardware and software. With this more flexible computational building block, along with reconfigurable compute slices and deterministic (rather than dynamic) execution on the silicon, TCP reduces energy-wasting data movement while maintaining the flexibility to support new AI models and new optimization techniques. Developers can load models from Hugging Face, define models in PyTorch, and use a drop-in replacement for vLLM. Developers can also drop down to Furiosa’s Tensor Contraction Language and a low-level virtual ISA when fine-grained optimizations are required. The result is more tokens from a fixed power budget. Furiosa’s 180-watt, air-cooled RNGD card delivers throughput comparable to a GPU drawing twice as much power or more.

View source version on businesswire.com: https://www.businesswire.com/news/home/20261028655033/en/