The emergence of DeepSeek, a Chinese startup, has sent shockwaves through the tech industry, particularly affecting Nvidia, the leading AI chipmaker. After DeepSeek launched its AI model DeepSeek-R1 on January 20, 2025, claims emerged stating it could perform comparably to models from U.S. giants like OpenAI and Meta, but at significantly reduced costs.
According to reports, DeepSeek-R1 reportedly delivers performance on par with OpenAI's renowned models but is 20 to 50 times cheaper to use, depending on the task. This efficiency has raised alarm bells within the technology sector, leading to fears for Nvidia’s market stability. On the trading day following DeepSeek's announcement, Nvidia’s shares plummeted more than 17%, resulting in CEO Jensen Huang losing approximately $18 billion from his net worth alone.
Nvidia's spokesperson acknowledged the competitive threat posed by DeepSeek, stating, "DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling." The company emphasized the need for significant Nvidia hardware to support DeepSeek's models, underlining the dependence on chips for training and operations.
What’s more intriguing is how DeepSeek managed to achieve these advancements. Reportedly utilizing around 50,000 Nvidia H100 chips, the startup’s models suggest significant training cost reductions. Alexandr Wang, the CEO of Scale AI, expressed doubts over DeepSeek's transparency during his media interactions, speculating about the startup's undisclosed quantities of Nvidia chips due to U.S. export controls.
DeepSeek's rapid rise can be seen as part of broader trends among Chinese tech firms intent on catching up to U.S. artificial intelligence capabilities. Following the successful launch of ChatGPT by OpenAI, the Chinese tech environment became electrified, sparking rapid developments among its domestic competitors. The hype around DeepSeek signifies not just technological capability but also potentially geopolitical ramifications, especially as relations between the U.S. and China grow increasingly fraught.
Despite initial skepticism, DeepSeek's models have begun gaining traction, evidenced by its AI Assistant surpassing ChatGPT as the top-rated free app on Apple's App Store. This shift not only highlights DeepSeek's impressive technological advancements but also reflects growing consumer enthusiasm for alternatives to established U.S. models.
Investors, initially reacting with panic to the news, may need to reassess the long-term impacts of these developments. While concerns around AI chip expenditures are valid, some analysts suggest demand for GPUs may rise due to the competitive nature of the market DeepSeek is helping to create.
DeepSeek is linked to High-Flyer, co-founded by controlling shareholder Liang Wenfeng, which has previously focused on quantitative trading. Their deep involvement with AI and chip technology places them strategically to innovate within the sector. Analysts have noted the potential for DeepSeek's research initiatives to shift the balance of power, especially if they can maintain the performance of their models without extensive financial input.
The world is now watching closely as DeepSeek not only develops its products but also navigates the complicated waters of global tech competition and U.S. export regulations. Geopolitical dimensions will also likely play a role, as China's ambitions to become self-sufficient across strategic industries confront U.S. attempts to limit technological advancements on the Chinese front.
Given these dynamics, Nvidia and U.S. tech giants must grapple not only with the immediate impact of DeepSeek's advancements on stock values and consumer sentiment but also with longer-term strategic shifts within the AI industry. The stakes are high, and the global tech order may be on the cusp of transformative change due to this new player.