Chinese artificial-intelligence start-up 01.AI bought enough of Nvidia’s (NASDAQ:NVDA) chips earlier this year to last for the next 18 months, before the U.S. brought in stricter export curbs for the sale of chips and chip tools to China, Bloomberg News reported.
Nvidia’s graphics processing units, or GPUs, are considered to be the most advanced for training AI models, such as OpenAI’s GPT4.
“We have stockpiled a lot of Nvidia chips,” said Kai-Fu Lee, during a Bloomberg Television interview, according to the report.
Lee, who is also CEO of venture capital fund Sinovation Ventures, founded 01.AI earlier this year. The start-up is valued at over $1B.
China’s President Xi Jinping has already called for hastening efforts to make breakthroughs in core technologies. A plan released by the Chinese government showed that China intends to step up efforts to boost its computing power by more than a third in less than three years, aimed at benefiting local suppliers and strengthening technology self-reliance, amid U.S. export curbs.
However, Lee noted that it is not clear if China can build the required technologies before companies like his run out of foreign supplies, according to the report.
“The jury is out on whether China in 1.5 years can make equivalent or nearly as good chips,” said Lee, who has worked at Apple (AAPL), Microsoft (MSFT) and Alphabet (GOOG) (GOOGL) subsidiary Google’s China unit.
Lee thinks that the tensions between the U.S. and China divide the technology world even more.
“Americans will supply their products and technologies to the US and other countries and Chinese companies will build for China and whoever else uses Chinese products. The reality is that they will not compete very much in the same marketplace,” Lee commented.
01.AI has an open-source large language model, or LLM, called Yi-34B, which according to the Chinese company outperforms much larger models like Meta Platforms’ (META) Llama 2.
Lee noted that developing AI models is very costly but there is no option as there is a direct co-relation between money spent on GPUs and performance of the products. 01.AI raised money from Alibaba’s (BABA) cloud unit and borrowed from Sinovation ahead of that to buy chips, according to the report.
Lee added that each time when money is raised, a major part, perhaps three quarters, is being put into buying GPUs.
Lee said in a separate interview that he expects significant revenue next year, but “next to zero” this year.
High-performing semiconductors are vital building blocks from AI to smartphones.
On Thursday, it was reported that Nvidia (NVDA) intends to announce three new chips for China, weeks after the company was restricted by the U.S. from selling two high-end AI chips and another top gaming chip to Chinese companies.
In October, the U.S. brought in new updates to its export restrictions which curbed the sale of chips that Nvidia (NVDA) made for the Chinese market, such as the A800 and H800 chips.
The move was part of Washington’s efforts to hinders China’s access to advanced semiconductor technology. In August, U.S. President Joe Biden had signed an executive order to regulate certain U.S. investments in China in three areas — semiconductors and microelectronics, quantum information technologies and AI.
Generative AI services have become the talk of the town since the launch of Microsoft-backed OpenAI’s ChatGPT last year. Alibaba’s (BABA) Tongyi Qianwen 2.0 and Tongyi Wanxiang, Baidu’s (BIDU) Ernie Bot, OpenAI’s text-to-image tool DALL·E 3, Meta Platforms’ (META) AudioCraft, SeamlessM4T, and Llama 2, Google’s Bard, Samsung’s (OTCPK:SSNLF) Gauss, and Getty Images’ (GETY) model called Generative AI by Getty Images, are some of the LLMs, among the many, being developed by companies worldwide.