Seoul-based AI chip startup FuriosaAI raised a $620M Series D at a $2.9B valuation, up from $735M in 2025; South Korea's National Growth Fund contributed $280M (Yoolim Lee/Bloomberg)

FuriosaAI, a Seoul-based AI chip startup, completed a $620 million Series D funding round. The investment valued the company at $2.9 billion. That is a major increase from its $735 million valuation in 2025. The round gives FuriosaAI substantially more capital to develop its business.

South Korea's National Growth Fund contributed $280 million. The source does not identify the other investors or explain how FuriosaAI will spend the money. In general, a funding round exchanges company ownership for new capital, which startups use to grow products, teams, operations, or sales.

The deal strengthens FuriosaAI's attempt to challenge Nvidia Corp., the dominant name identified in the article's comparison. Its much higher valuation signals stronger investor confidence, but it does not guarantee commercial success. FuriosaAI must still turn the funding into competitive chips and business growth.

A Series D round is typically a startup's fourth major institutional equity funding stage, following Series A, B, and C. It usually involves a more mature company with a product, customers, or meaningful growth. Investors provide capital in exchange for shares, using the company's latest valuation to price the deal.

Startups commonly use Series D money to increase production, hire specialists, improve technology, expand sales, enter new markets, or prepare for a later public offering. The exact use depends on the company's needs. The source does not state FuriosaAI's specific spending plan, so these are general uses rather than confirmed plans.

FuriosaAI raised $620 million at a $2.9 billion valuation. That size suggests an ambitious growth phase. The money could help the Seoul-based company build scale and strengthen its effort to challenge Nvidia, but funding alone does not establish market leadership.

FuriosaAI's valuation increased by $2.165 billion, rising from $735 million in 2025 to $2.9 billion in the new funding round. Dividing $2.9 billion by $735 million gives roughly 3.95, so the company's value almost quadrupled. The article describes the increase as quadrupling.

A company valuation is the estimated worth assigned to the business by investors during a financing deal. It can reflect expectations about technology, growth, revenue, market opportunity, and competitive strength. It is not necessarily cash sitting in the company's bank account, and it can change in a later transaction.

For FuriosaAI, the higher valuation means investors accepted a much larger price for ownership than in 2025. It marks stronger financial backing and expectations as the company challenges Nvidia. Still, valuation is an assessment, not proof that FuriosaAI has already achieved comparable scale or profits.

The National Growth Fund is identified in the article as South Korea's government-backed investment fund. The source gives no further details about its structure, objectives, or investment rules. In general, such funds invest public or state-supported capital in businesses considered important to national economic growth.

Its $280 million contribution went to FuriosaAI, a Seoul-based AI chip startup. AI chips are strategically important because they support the computing systems behind modern artificial intelligence. Investing in a domestic chip company can provide capital for research, engineering, manufacturing, and expansion. Those possible uses are general, not confirmed plans for this round.

The investment also gives FuriosaAI substantial backing while it tries to challenge Nvidia. The article does not state the fund's exact reason for investing or whether it expects a financial, industrial, or national-security return. Its participation shows that FuriosaAI's financing includes significant support from a South Korean government-backed investor.

The new funding gives FuriosaAI resources to accelerate its business. The article does not say exactly where the money will go, so possible uses include chip research, engineering talent, software development, manufacturing capacity, customer support, and sales. These activities can help an AI-chip startup turn technology into a reliable commercial product.

For example, FuriosaAI could invest in chips that process AI workloads more efficiently, then support those chips with software tools that make them easier to use. It could also increase production and work with more customers. Better performance, lower operating costs, and dependable supply would make its products more attractive compared with established alternatives.

Nvidia has a major position in the AI-chip market, and the article frames FuriosaAI as a challenger. The funding improves FuriosaAI's ability to compete, but it does not guarantee success. The company must convert capital into products, adoption, and scale.

An AI chip is a processor built or optimized for workloads such as training and running machine-learning models. These workloads involve enormous numbers of repeated mathematical operations, especially on matrices and vectors. AI chips can perform many of those operations simultaneously, reducing the time and energy required.

Ordinary computer processors, called CPUs, are general-purpose devices. They handle many kinds of tasks and are flexible, but they usually contain fewer highly parallel computing units. AI accelerators, including GPUs and specialized neural processors, devote more hardware to parallel calculations and often include fast memory systems for moving model data.

That difference matters because AI models can require substantial computing power. Specialized chips may deliver better performance per dollar or per watt for those workloads, while CPUs remain useful for general tasks. FuriosaAI develops AI chips, placing it in a market where efficient hardware can determine how quickly and affordably companies build and use AI systems.

AI models rely on huge numbers of repeated calculations during training and use. Specialized chips matter because they can perform many of these operations in parallel and move model data efficiently. That can shorten training time and make responses cheaper or faster when models run in products. Energy efficiency also matters because AI systems use large data centers.

Nvidia became the market leader by combining powerful parallel-processing hardware with software tools, developer support, and a broad commercial ecosystem. Its scale can help fund research, manufacturing relationships, and customer support. Compatibility across its products can also reduce the effort required to move AI applications from development into production.

These advantages create a difficult target for FuriosaAI and other challengers. FuriosaAI's $620 million Series D gives it more resources to develop and commercialize competing chips. However, it must win on more than raw chip performance. Software, reliability, availability, price, and customer adoption will also shape the contest.