Nvidia, the world’s largest chipmaker, has apparently discovered a means to sell high-end chips to Chinese companies while keeping compliant with US restrictions aimed at limiting China’s access to technology.
- IMPORTANT NOTES
- Nvidia, the world’s largest chipmaker, has purportedly discovered a means to sell high-end chips to Chinese corporations while keeping compliant with US rules.
- According to insiders, Nvidia will supply three new chips to domestic manufacturers in the next days, according to Chinese financial publication Cailian Press.
- Nvidia, the US Department of Commerce, and the Bureau of Industry and Security did not react promptly to a request.
China accounts for 20% to 25% of Nvidia’s sales in its largest sector, data centers. According to insiders, It will supply three new chips to domestic manufacturers in the next few days, according to Chinese financial publication Cailian Press. According to the article, the chips are dubbed HGX H20, L20 PCle, and L2 PCle and are based on Nvidia’s H100 chip.
The H100 and A100 artificial intelligence processors were the first to be affected by new US restrictions aimed at limiting shipments to China last year. In a September 2022 filing, It stated that the US government would still allow it to develop the H100 in China.
“In the near term, Chinese manufacturers have no better option, and they will continue to buy Nvidia’s chips while searching for replacements.” -Bo Du(MANAGING DIRECTOR, WESTSUMMIT CAPITAL MANAGEMENT)
Companies in China then shifted to Nvidia’s H800 and A800 processors, although the United States recently imposed new limitations on those shipments. According to Bo Du, managing director at West Summit Capital Management and a veteran chip engineer, the H20′s computing capability is just around half that of the A100.
That amounts to “basically saying goodbye to physical simulation,” he added in Mandarin, according to Reports. While it is conceivable to support massive model simulations with clusters of lower-power CPUs, he claims that there is no optimum option given the expense.
“In the short term, Chinese manufacturers have no better option and will continue to buy Nvidia’s chips while looking for replacements,” Du said, saying that several significant internet companies have begun to purchase domestically-made AI chips on a wide scale.
As Chinese enterprises hurry to construct local versions of OpenAI’s ChatGPT, demand for artificial intelligence computer capacity has only increased. The Financial Times also reported about it’s new processors for the Chinese market, citing a document sent by the chipmaker to potential customers.
It did not respond to requests for comment. The US Department of Commerce and the Bureau of Industry and Security did not reply quickly to a request. According to a Thursday online post by research firm SemiAnalysis, all three of Nvidia’s new processors have operating parameters that fall beyond the scope of US limitations. The company claims to have over 64,000 subscribers to its Substack tech newsletter.
“Nvidia is perfectly straddling the line on peak performance and performance density with these new chips to get them through the new US regulations,” claimed SemiAnalysis in a press release. Nomura analysts previously discovered that Nvidia’s Drive AGX Orin processor did not match all of the conditions for a US limitation on sales to China, allowing Chinese electric car makers to continue using the technology.
The United States has stated that its priority is to limit China’s development of advanced technology for military purposes. The government of President Joe Biden has also stressed the country’s battle with China. Domestic players are attempting to devise loopholes for US prohibitions.
Huawei unveiled a smartphone in late August that, according to evaluations, offers download speeds comparable to 5G, thanks to an upgraded semiconductor chip. It’s unclear whether the current chip fabrication used older equipment or different buying methods.
Frequently asked questions
What is going on at NVIDIA?
Amdocs with NVIDIA to Accelerate Generative AI Adoption in the $1.7 Trillion Telecom Industry. Amdocs and NVIDIA announced today that they will work together to optimize large language models (LLMs) in order to accelerate the adoption of generative AI products and services across the $1.7 trillion telecommunications and media industries.
Is NVIDIA making money?
It forecasted third-quarter sales of $16 billion, a 170% increase. Analysts now estimate it’s earnings to increase by 222% this year, while sales will increase by 104%. Nvidia’s earnings per share plummeted 25% last year.
Who is the CEO of Nvidia?
Jensen Huang created NVIDIA in 1993 and has served as president, chief executive officer, and member of the board of directors since then.
Click here, to check out the latest post on Instagram.
Also Read: Emirates Group: Emirates Group Reports A $2.7 Billion Half-Year Profit
image source: google
