In a move that has left the tech world reeling, a Chinese startup DeepSeek has shaken the foundations of the artificial intelligence (AI) industry, igniting a massive selloff that wiped $1 trillion from global stock markets.
The startup’s introduction of an open-source, cost-effective large language model (LLM) has sent tremors through the AI ecosystem, including tech heavyweights like Nvidia, Broadcom, and TSMC. Investors, rattled by the news, quickly adjusted their portfolios, fearing that the AI arms race might no longer require the vast, expensive infrastructure once thought necessary.
Key facts
- DeepSeek, a China-based startup, unveiled its R1 LLM model, which it claims matches or outperforms existing AI models at a fraction of the cost.
- DeepSeek’s model was reportedly trained for under $6 million using midrange Nvidia H800 chips, challenging the notion that AI breakthroughs require massive investments.
- The news sent shockwaves through the Nasdaq, erasing nearly $1 trillion from the market cap, particularly affecting AI-adjacent stocks like Nvidia and TSMC.
- Nvidia saw its market value drop by a historic $600 billion, the biggest drop for any company on a single day in U.S. history.
- The drop knocked Nvidia from its position as the world’s most valuable company, sending its valuation from $3.5 trillion to $2.9 trillion, less than Apple’s and Microsoft’s.
- Despite the panic, Nvidia downplayed the threat, calling DeepSeek’s LLM an “excellent AI advancement.”
DeepSeek Shockwave
DeepSeek, a Chinese startup founded in 2023, has unveiled its R1 model—a large language model that rivals the best from OpenAI and Anthropic and is also open-source and free to use.
The implications were immediate and severe. Investors, fearing that the AI spending boom might be rendered obsolete by cheaper, more efficient alternatives, began offloading shares of tech giants.
The Nasdaq 100, the tech-heavy index, fell 3%, its steepest decline in six weeks. Nvidia, whose high-end GPUs are the backbone of AI development, saw its market value plummet by $600 billion on Monday, January 27, in the largest single-day loss in U.S. history.
Other AI-adjacent stocks, including Broadcom and TSMC, also took significant hits. Even the “Magnificent 7” stocks—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla—were not spared, with premarket trading showing sharp declines across the board.
Is there really a cheaper, faster path to AI?
What makes DeepSeek’s R1 model so disruptive is its cost efficiency. According to the company, the model was developed in just two months at a cost of under $6 million, using Nvidia’s lower-performing H800. This is a big difference from the billions spent by U.S. tech giants like OpenAI and Meta on their LLMs.
Despite being built on mid-range Nvidia H800 chips—considered less powerful than the latest hardware typically used for such projects—DeepSeek’s model reportedly matches or even outperforms industry-leading systems like OpenAI’s GPT and Meta’s Llama.
DeepSeek-R1 is 20 to 50 times cheaper to use than OpenAI’s o1 model, depending on the task, according to a post on DeepSeek’s official WeChat account.
Nvidia, which has become a critical player in AI development thanks to its high-performance GPUs, moved quickly to downplay the threat from DeepSeek. The company acknowledged the significance of the new model but framed it as a testament to the potential of AI’s scalability, calling DeepSeek an “excellent AI advancement” rather than a direct challenge to its dominance. “DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant,” said an Nvidia spokesperson.
AI’s Sputnik moment?
Silicon Valley venture capitalist Marc Andreessen called the DeepSeek’s R1 model AI’s “Sputnik moment”—referring DeepSeek’s emergence to the Soviet Union’s launch of Sputnik that stunned the U.S. during the Cold War and marked a significant shift in global technological leadership. “DeepSeek R1 is one of the most amazing and impressive breakthroughs I’ve ever seen — and as open source, a profound gift to the world,” he said in a separate post.
Other experts have raised concerns about the vulnerabilities in the U.S. tech sector’s dominance over AI. “We don’t know whether this is the ‘Sputnik Moment’ for stocks, but this is certainly a wake up call that we are not the only game in town,” Paul Nolte, market strategist and senior wealth manager at Murphy & Sylvest Wealth Management, said. “That requires a lot of investors to look at the AI companies in a different way: To put these very high valuations in the stocks thinking they have cornered the market is a huge mistake and that is being re-rated.”
What next?
By creating an AI model that rivals the capabilities of the most advanced U.S. systems at a fraction of the cost, DeepSeek may have achieved what was once considered impossible: a cheaper, more efficient AI model that could signal a new era in AI development.
As the industry digests the potential consequences of DeepSeek’s innovation, all eyes are now on the next round of earnings reports from AI-focused companies. For now, the market is showing signs of recovery, with Nvidia, TSMC, and Broadcom clawing back some of their losses. However, one thing is clear: The race for AI supremacy is far from over, and the stakes have never been higher.
1 comment
Comments are closed.