(Bloomberg) -- British chip designer Graphcore Ltd. is laying off most of its staff and discontinuing sales in China, marking another setback for a startup once heralded as a potential rival to Nvidia Corp.
The company confirmed the decision, citing recent US export controls that restrict the sale of technology into China. “Regrettably, this means we will be significantly scaling back business operations in China,” a spokesperson said in an email. The company declined to share the number of employees impacted.
Graphcore, which formed in 2016, designs semiconductors tailored to support artificial intelligence software. Investors poured money into the startup as they searched for a viable alternative to Nvidia, whose gear is in high demand. In 2020, Graphcore raised $222 million at a $2.8 billion valuation, making it one of the UK’s most promising startups.
But the Bristol, England-based company has struggled to get commercial traction. Revenue fell 46% and losses widened 11% to $204.6 million in 2022, according to the most recent company filing. In October, it disclosed that it needed to raise funds to keep operating. Graphcore hasn’t announced any funding since.
Venture capital firm Sequoia, one of Graphcore’s most notable backers, wrote down the value of its stake in the startup to zero, the Sunday Times reported earlier this year.
Nigel Toon, Graphcore’s chief executive officer, had previously pointed to China as a potential growth market, particularly since trade war restrictions have hampered Nvidia’s ability to sell into the country. At the Bloomberg Technology conference in London in October, Toon said sales from China accounted for “maybe 20 to 25%” of his company’s business.
“Elsewhere, the need for AI compute continues to increase and Graphcore is working with customers around the world to meet their demand for a powerful, cost-effective alternative to GPUs,” the Graphcore spokesperson said, referring to the powerful graphics chips that Nvidia and others make.
The US export controls have focused primarily on limiting China’s access to high-end chips and other equipment used for AI.
©2023 Bloomberg L.P.