Why DeepSeek Is Making a Major Impact on Tech Stocks, Such as Nvidia


**DeepSeek’s Disruption: A Rough Day for the Stock Market and AI Sector**

Today could prove to be a challenging day for the stock market, with the burgeoning Chinese AI startup DeepSeek potentially at the forefront of these developments.

DeepSeek has been generating significant buzz in the technology sector following the debut of its newest reasoning AI model, [DeepSeek R1](https://mashable.com/article/deepseek-ai-hands-on). This model is not only competitive—if not superior—to alternatives like OpenAI’s main offerings, but it is also available at an incredibly low cost. Users can access DeepSeek R1 for free through web and mobile apps, and its open-source framework allows for local operation, customization, and alleviates worries regarding privacy or Chinese regulatory constraints.

Within a matter of days, DeepSeek’s traction skyrocketed, driving it to the [top of Apple’s App Store](https://mashable.com/article/deepseek-chatgpt-apple-app-store), where it overtook OpenAI’s ChatGPT as the most downloaded application. Nevertheless, while this achievement has delighted AI supporters, it has caused significant ripples in the stock market, especially affecting Nvidia investors.

### Nvidia Stock Suffers

In early Monday trading, Nvidia’s shares plummeted by as much as 13% in premarket trading, threatening to wipe out over $500 billion from the company’s market capitalization. As it stands, Nvidia’s stock ($NVDA) is priced at $125.95, reflecting an approximate 11% decline from its last closing value of $142. Concurrently, the NASDAQ has dropped 3.9% in premarket trading.

The market downturn is attributed to DeepSeek’s economical and efficient performance. Although information about DeepSeek’s financials is sparse, it has been reported that the company spent only [$5.58 million](https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf) to train its DeepSeek V3 model. The company is thought to have utilized 50,000 Nvidia GPUs—a substantial figure, yet still considerably fewer than the resources wielded by leading U.S. AI firms, which reportedly have access to at least ten times that number.

### Industry Titans Take Note

DeepSeek’s ascent has captured the attention of significant names in the industry. During his address at the World Economic Forum in Davos, Switzerland, Microsoft CEO Satya Nadella highlighted the importance of DeepSeek’s innovations.

“To observe the DeepSeek new model, it’s incredibly impressive in terms of how they have adeptly accomplished an open-source model that does this inference-time compute while being super-compute efficient,” [Nadella stated](https://www.cnbc.com/2025/01/24/how-chinas-new-ai-model-deepseek-is-threatening-us-dominance.html).

In a similar vein, renowned investor Marc Andreessen [enthusiastically endorsed](https://x.com/pmarca/status/1882719769851474108) DeepSeek as “one of the most remarkable and impressive breakthroughs” he has ever encountered.

### Industry Impact

Should DeepSeek have successfully matched OpenAI’s leading models while utilizing notably fewer Nvidia GPUs and at a significantly lower price point, the ramifications for the AI sector could be extensive. This situation may indicate a potential downturn in demand for Nvidia GPUs, which have been a critical factor in the company’s meteoric stock value. On a broader scale, it raises doubts about whether the existing AI market is overhyped, as reflected in today’s [performance of AI-related stocks](https://stoxx.com/index/ixagal1p/).

Nevertheless, the landscape is complex. DeepSeek, a spin-off from the Chinese AI-oriented hedge fund High-Flyer Quant, might be operating with more GPUs and a greater budget than what is publicly acknowledged. Moreover, while the initial market feedback has been unfavorable, DeepSeek’s technology could unveil fresh avenues in the AI domain, potentially reigniting competition in AI hardware down the line.

### Competition Intensifies

DeepSeek is not the sole Chinese entity making waves in the AI arena. ByteDance, the parent organization of TikTok, recently unveiled its [Doubao-1.5-pro model](https://www.reddit.com/r/LocalLLaMA/comments/1i9wbya/bytedance_announces_doubao15pro/), which outshines numerous leading models across various benchmarks. However, unlike DeepSeek, Doubao lacks an open-source model, which might restrict its accessibility and attractiveness.

### The Future Path

It will require time for the market to thoroughly assess the consequences of DeepSeek’s emergence and what it could mean for the AI industry. Presently, sentiment is notably negative, with stock values trending downward. Whether this signals the start of a significant transformation in the AI landscape or is merely a fleeting market correction remains undetermined.

For both investors and tech aficionados, one fact stands out: DeepSeek has made its mark, and