Sapeon, a South Korean artificial intelligence chip startup headquartered in California, is raising a funding round that puts its valuation above $400 million, its CEO told CNBC.
The startup is backed by massive South Korean firms SK Telecom, memory-chip maker SK Hynix and SK Square, an investment company spun off of SK Telecom.
Sapeon designs artificial intelligence semiconductors for data centers, for example those run by cloud computing firms. These AI chips are required for AI applications that require huge amounts of data processing.
Currently, U.S. firm Nvidia
dominates this market. But a number of established players such as AMD
and startups like Sapeon, are looking to challenge Nvidia.
“AI solutions will grow a lot thanks to the evolution of AI services like ChatGPT,” Soojung Ryu, CEO of Sapeon, told CNBC in an interview at the Mobile World Congress that aired Wednesday.
ChatGPT is the viral chatbot developed by OpenAI. It has caused a stir in the tech world, with giants from Google to Chinese firm Baidu scrambling to release their own rivals. Technology executives say ChatGPT has put AI applications on the map.
Analysts at Bernstein expect ChatGPT to be a multi-billion dollar boon for chipmakers looking to fuel these AI models.
“We would like build this kind of system (AI chips) to have an opportunity for the business,” Ryu said.
Sapeon was created in 2016 inside SK Telecom, one of South Korea’s biggest telecommunications firms. SK Telecom then spun out the firm last year and raised outside investment.
Ryu said the company is currently raising money at a valuation of $400 million.
South Korean semiconductor firms including Samsung and SK Hynix have typically been strong in memory chips that go into devices like PCs. Sapeon is among one of the first firms out of South Korea to try and crack the AI chip market.
Ryu said that Sapeon is targeting the U.S. market which would pit it against Nvidia. When asked if Sapeon can challenge Nvidia, Ryu answered “yes.”
Currently, Sapeon has one chip on the market called the X220. It is built on so-called 28-nanometer technology. The nanometer figure refers to the size of each individual transistor on a chip. The smaller the transistor, the more of them can be packed onto a single semiconductor. Typically, a reduction in nanometer size can yield more powerful and efficient chips.
Ryu said the company will launch a 7-nanometer AI chip this year and it will be manufactured by TSMC, the world’s biggest contract chipmaker. That would bring it closer to the current technology on the market.