FuriosaAI raises 820 billion won in funding to scale AI chip production

Record-breaking funding round

FuriosaAI recently completed a Series D funding round that marks a significant milestone for the South Korean technology sector. The company raised 820 billion won, which represents the largest funding round to date for a domestic AI chip startup. This influx of capital places the company's valuation at approximately 3.8 trillion won.

The funding process began in January with an initial target of 450 billion to 750 billion won. However, the company reported that the final amount increased significantly due to high demand from investors. This surge in interest highlights the growing global appetite for specialized AI semiconductor hardware.

The National Growth Fund served as the primary investor, contributing 370 billion won to the round. The funding also drew in a diverse group of international and domestic players, including U.S.-based Infinitum, Singapore's QUDA Capital Partners, and Cisco Investments. The participation of these major entities suggests a high level of confidence in FuriosaAI's technical roadmap.

Expanding the production line

A primary goal for the new capital is the expansion of production and supply capacity for the RNGD chip. Known as Renegade, this second-generation AI chip entered the mass production phase in January. The company intends to use the 820 billion won to scale this output to meet increasing global demand.

FuriosaAI is not just focusing on manufacturing volume but also on the reach of its technology. The company plans to use the funds to acquire new global customers and strengthen its existing software ecosystem. By building a robust software layer around the hardware, they aim to make their chips more accessible to developers.

The company is already seeing early adoption of the RNGD chip in various commercial applications. For example, Samsung SDS has launched an NPUaaS, which is a service providing access to Neural Processing Units (NPUs), based on the RNGD chip. Additionally, the chip will be supplied to Upstage for its next search summarization service.

Strategic corporate partnerships

The funding round included significant participation from established tech giants and financial institutions. Naver, DSC Investment, and KDB Capital all participated as follow-on investors, while the Export-Import Bank of Korea made its first-ever direct investment in a venture company. These partnerships provide FuriosaAI with a stable foundation of institutional support.

The company is also working closely with LG AI Research to verify technology and cooperate on commercialization efforts. These collaborations are essential for ensuring that the hardware meets the rigorous standards required by enterprise-level AI applications. By working with established names like LG, FuriosaAI can validate its technology in real-world environments.

The investment from Infinitum is particularly notable given the firm's history in the semiconductor space. Infinitum previously co-led a $650 million investment round in the AI semiconductor company Groq in June. This suggests that the investors are actively seeking to diversify the global AI infrastructure landscape with high-performance hardware options.

Roadmap for global infrastructure

Looking ahead, FuriosaAI is aiming to position itself as a major player in the global AI infrastructure market. The company is currently developing a third-generation AI accelerator in collaboration with Broadcom. This partnership is intended to accelerate the development of products specifically designed for hyperscale AI infrastructure.

Hyperscale infrastructure refers to the massive computing resources used by large cloud providers to train and run large-scale AI models. By targeting this segment, FuriosaAI seeks to compete directly with the dominant hardware providers currently leading the market. The collaboration with Broadcom provides the technical expertise needed to reach this scale.

CEO June Paik stated that the company can now begin to spread its technology to the global AI infrastructure market. The company's strategy focuses on providing a new option for the market based on high-performance and high-efficiency semiconductors to meet the needs of global AI users.