There is more bad news coming for the US chip giant Nvidia from China. According to a report in Reuters, Chinese startup DeepSeek is developing its own AI chip. This means that the hardware push from one of China’s biggest AI stars could reduce its reliance on Nvidia. Not to forget that China also has Huawei making AI chips that Nvidia CEO Jensen Huang has openly warned about. As per the report, quoting sources, Deepseek’s chip is designed for inference — the stage of AI computing in which a trained model generates responses for users — rather than for training new models.DeepSeek’s inference chip is expected to target the fastest-growing segment of AI computing demand. As AI applications spread, more of the industry’s computing work is shifting from training models to running them, which relies on specialised chips that can be cheaper and less power-hungry than general-purpose GPUs. If successful, DeepSeek’s expansion into semiconductor development would mark a major strategic shift for a company widely hailed in China as the country’s AI champion. Incidentally, this will not only make things tough for Nvidia but also potentially add to challenges for Chinese tech giant Huawei. For DeepSeek, the effort carries an added strategic dimension. As US export controls bar Chinese companies from buying Nvidia’s most advanced chips, Beijing has been pressing its technology champions to build domestic alternatives.
Nvidia’s AI chip sales in China have almost stalled
Nvidia is one of the biggest victims of the ongoing race between the US and China to lead artificial intelligence space. The tensions have heated up the battle over hardware and computing power between the two countries. Controls imposed by Washington on exports of advanced technology due to national security concerns initially stalled sales of Nvidia’s advanced H200 AI chips there. By the time Huang won a reprieve, with Trump agreeing to their sale, Beijing had switched to encouraging use of domestically designed chips made by local rivals led by Huawei.“Well, we were in China for 30 years, and before the export control banned Nvidia out of China we had about 95% market share, and so we were competing just fine,” he said in a recent interview with The Associated Press. “We have to have, number one, make sure that we have national security and that we protect our nation, but we also simultaneously should go and compete and grow our technology industry and maximize our exports,” he said.DeepSeek’s model that wiped hundreds of billions from Wall Street is based on Nvidia chipsOn its part, DeepSeek has used both Nvidia and Huawei chips. The company has said the foundation model underpinning R1, the reasoning model whose low-cost performance that triggered a complete rout in US tech stocks in January 2025, was trained on Nvidia’s H800, a chip designed for the Chinese market that Washington banned in late 2023.The company has since leaned increasingly on its home country Nvidia rival — Huawei. In April this year, DeepSeek released its V4 model adapted for Huawei’s Ascend chips, and Huawei said its processors were used in part of the training of V4-Flash, a lighter version of the model. Orders for Huawei’s Ascend 950 chips from Chinese tech conglomerates surged after the launch, Reuters has reported.
DeepSeek joins OpenAI and Anthropic with its in-house chips
With an in-house chip, DeepSeek would be joining other global AI developers in seeking greater control over the hardware behind their models and reduced dependence on Nvidia’s. OpenAI last month unveiled Jalapeno, its first custom inference chip, developed with Broadcom, while Anthropic has been weighing building its own AI chips.For now, DeepSeek’s effort to join the AI chip race reportedly remains at an early stage, with the company reaching out to external partners and holding discussions with chip-design, foundry and memory companies. The Hangzhou-based company has also increased hiring of chip-design engineers in recent months, but recruitment has reportedly been done privately without job postings on public hiring platforms.