China's Deepseek Plan To Make Own AI Chips Causes Nasdaq Sell-Off

As if the US AI industry weren't already nervous enough as skepticism about the overhype grow, the NASDAQ fell sharply yesterday after Chinese AI leader Deepseek announced it would be making its own AI inference chips. This is - correctly - perceived as a direct threat to Nvidia's dominance and, by extension, to US AI hegemony. 

Deepseek's Chinese frenemies, Alibaba and Baidu, are contemplating similar moves. They are being driven, in part, by US chip export controls, but also by a desire to capture more of the AI stack margins for themselves. This is yet another example of how, between token costs and chip hoarding, getting too greedy could cost the AI industry and the US. JL

Samuel Axon reports in ars technica and Shashwat Chauhan reports in Reuters:

Deepseek, the Chinese startup developing large language models competitive with those from OpenAI and Anthropic, is planning to enter the silicon business. The focus is on data center chips for inference, not training, and the goal is to reduce reliance on both Huawei and Nvidia.  The Nasdaq fell more than 1%  ⁠Tuesday, pressured by a selloff in semiconductor stocks ⁠including Nvidia amid mounting doubts about the sustainability of the AI-driven rally, while ​DeepSeek making an AI chip also soured sentiment. Chinese tech giants Alibaba and Baidu have been making such moves, too. While chip export controls in the US are a major reason this is an urgent concern for DeepSeek, US-based AI companies are making similar chip plans. OpenAI’s is a play to reduce its reliance on Nvidia, but also to have Apple-like control over the entire tech stack for its products.

DeepSeek, the Chinese startup developing large language models that are competitive with those from US companies like OpenAI and Anthropic, is planning to enter the silicon business.

Citing three people familiar with the matter, Reuters writes that DeepSeek has been working on a move into silicon for about a year. It has been meeting with potential partners in the hardware and silicon space and has been hiring engineers for the project.

The focus is on data center chips for inference, not training, and the goal is likely to reduce reliance on both Huawei and Nvidia.  

The Nasdaq fell more than 1% ⁠on ⁠Tuesday, pressured by a selloff in semiconductor stocks ⁠including Nvidia amid mounting doubts about the sustainability of the AI-driven rally, while a report on ​China's DeepSeek making an AI chip also soured sentiment.

Nvidia shed 0.7% after Reuters reported Chinese startup DeepSeek is developing its own AI chip, a push ‌that could reduce its dependence on Nvidia and ‌Huawei chips.

Chip stocks tumbled across the world despite memory chip giant Samsung Electronics reporting a 19-fold jump in second-quarter operating profit and surpassing ⁠its combined earnings ⁠over the past three years.

Nvidia is the chipmaker for most AI companies in North America and Europe, but a United States export ban has prevented the company from achieving a similar presence in China. Huawei controls about half of the data center chip market there, and DeepSeek isn’t the only one trying to enter; Chinese tech giants like Alibaba and Baidu have been making moves, too.

While chip export controls in the US are a major reason this is an urgent concern for DeepSeek, US-based AI companies are making similar chip plans.

For example, OpenAI and Broadcom jointly announced Jalapeño, the former’s first chip designed for inference at scale, just a couple of weeks ago. Anthropic, too, has been exploring custom chip design, though there have not been any publicly visible milestones yet. 

In OpenAI’s case, it’s partly a play to reduce its reliance on Nvidia, but it’s also a desire to have Apple-like control over the entire tech stack for its products. Further, getting in at the silicon and data center levels can be an advantage in a market where data center access is likely to remain constrained, with multiple companies competing for compute as they scale up their AI models and services.