China's DeepSeek R1 AI Model Hits Tech Companies Like NVIDIA Shares



Tech stocks involved in artificial intelligence (AI) development around the world took a hit yesterday after a China-based company launched its open-source AI model, DeepSeek R1. DeepSeek R1 was found to outperform several models developed by OpenAI and Meta.


What shook the industry was that R1 was trained for just two months at a cost of around $6 million and used NVIDIA H800 GPUs that were depowered so that it could be sold to China by NVIDIA.


NVIDIA saw its shares fall by -16.97%, Broadcom -17.04%, TSMC -13.3%, Alphabet -4.04% and Microsoft -2.14%. China's success in producing high-performance AI models at a cost thousands of times lower dealt a major blow to the US AI industry, which has been using the motto that more powerful processing power is needed to develop better models.


OpenAI recently announced the Stargate Project, an investment of $500 billion over 4 years to develop AI data centers in the United States. With this project, OpenAI wants to ensure that the United States continues to be the world leader in AI technology.

Previous Post Next Post

Contact Form