In This Article:
Updated at 12:44 PM EST
Nvidia shares are on pace for their worst day in nearly five years in mid-day Monday trading following the launch of an AI-powered chatbot in China that has triggered a massive selloff in U.S. tech stocks and challenged the investment narrative tied to the world's hottest technology.
Nvidia (NVDA) shares are on pace to shed nearly $400 billion in market value, based on premarket declines. The market move follows the weekend revelation that DeepSeek, a China-backed startup, overtook OpenAI's ChatGPT as the world's most-popular AI tool on the Apple (AAPL) App Store.
💵💰Don't miss the move: Subscribe to TheStreet's free daily newsletter 💰💵
DeepSeek claims it was able to create and train a large-language AI model, using lower-end H800 Nvidia chips and open-sourced modeling, for less than $6 million. It also claims that its v3 LLM outperforms OpenAI, which cost around $540 million.
Llama, the open-sourced AI model developed by Meta Platforms (META) , reportedly cost around $500 million.
Related: Veteran trader overhauls Palantir stock price target after analyst review
If true, the DeepSeek revelation could test investors' patience for the capital-spending plans of the biggest hyperscalers, such as Microsoft (MSFT) , Amazon (AMZN) , Google parent Alphabet (GOOGL) and Facebook parent Meta, which are forecast to lay out as much as $300 billion this year alone.
That in turn could hurt the near-term profit and revenue forecasts for Nvidia, whose chips are the biggest component of the training and development of hyperscaler projects.
Nvidia's profit margins are currently holding in the low-to-mid 70% range, and its high-end Blackwell GPUs are the AI-sector benchmark for LLM training and inferencing.
DeepSeek threat to Nvidia sales?
The group is expected to generate around $197.3 billion in revenue in the coming fiscal year, which ends in January 2026, with net income of around $102 billion.
Citigroup analyst Atif Malik, however, isn't ready to concede that the DeepSeek launch represents a major threat to Nvidia's prospects, or its place at the epicenter of the AI investment thesis.
"While DeepSeek’s achievement could be groundbreaking, we question the notion that its feats were done without the use of advanced GPUs to fine tune it and/or build the underlying model through the Distillation technique," he said.
Related: Analyst reworks Nvidia stock price target on Blackwell demand forecast
That refers to a condition in which as smaller "student" AI model is trained to mimic a larger one at a fraction of the cost and with more speed and efficiency.