TLDR:
- Nvidia stock dropped 17% ($600B market value loss) after Chinese startup DeepSeek claimed to train an AI model for under $6M, compared to billions spent by U.S. companies
- Experts question DeepSeek’s cost claims, with Semianalysis estimating actual spending over $500M and uncertainty about which Nvidia chips were used
- Meta Platforms plans to increase AI capital expenditure by 66% to $65B in 2025, viewing heavy AI investment as a strategic advantage
- Microsoft CEO predicts exponential AI demand growth as technology becomes more efficient and accessible
- Nvidia integrated DeepSeek’s R1 model into its Enterprise software platform, potentially boosting sales of its premium GPUs
Nvidia’s stock experienced its largest single-day market value loss of nearly $600 billion after Chinese startup DeepSeek announced it had trained an AI model for less than $6 million. The market reaction highlighted investor concerns about potential disruption to Nvidia’s premium GPU pricing strategy.
DeepSeek claimed its R1 model, trained over two months, rivals OpenAI’s performance while using only a fraction of the traditional AI training budget. This contrasts sharply with U.S. tech companies’ multibillion-dollar investments in AI infrastructure.
However, industry experts have questioned DeepSeek’s cost claims. Consulting firm Semianalysis estimates the actual spending exceeded $500 million when accounting for all development stages. Questions also remain about which specific Nvidia chips DeepSeek utilized, particularly regarding potential use of more powerful GPUs acquired before U.S. export controls.

Nvidia currently holds approximately 80% of the AI chip market, with customers willing to pay premium prices for superior performance. The company’s chief, Jensen Huang, has maintained that Nvidia’s chips provide better value over time through reduced total ownership costs.
Meta Platforms has reinforced this premium strategy by announcing plans to increase its AI capital expenditure by 66% to $65 billion in 2025. CEO Mark Zuckerberg emphasized that the ability to invest heavily in AI infrastructure represents a strategic advantage.
Microsoft CEO Satya Nadella provided additional support for Nvidia’s position, noting that while AI training efficiency continues to improve, demand increases exponentially as the technology becomes more accessible. He referenced the Jevons paradox, suggesting that improved cost efficiency actually drives higher overall spending.
In response to DeepSeek’s announcement, Nvidia has integrated the R1 model into its Enterprise software platform as a microservice. This move demonstrates how Nvidia can leverage emerging AI models to drive demand for its premium GPUs, particularly for inference workloads.
Morgan Stanley analysts have revised their projections upward for AI infrastructure spending among major tech companies. The four largest hyperscalers – Amazon, Alphabet, Meta, and Microsoft – are now expected to increase spending by 32% to $317 billion in 2025.
Wall Street analysts maintain a median target price of $175 for Nvidia stock, suggesting 45% upside potential from its current price of $120. This indicates continued confidence in Nvidia’s market position despite the recent volatility.
The company’s strategic response includes emphasizing the importance of high-powered tools for AI inferencing, stating that significant numbers of Nvidia GPUs and high-performance networking remain essential for optimal results.
Technical experts note that achieving top AI model performance requires substantial computing power, regardless of initial training costs. This suggests that companies seeking competitive advantages will likely continue investing in premium GPU technology.
DeepSeek has acknowledged using Nvidia chips designed for the Chinese market, which are less powerful than the company’s main product line due to U.S. export controls. This detail adds complexity to comparisons between training costs and performance metrics.
Market observers point out that the initial stock reaction may have overlooked the distinction between training costs and ongoing operational requirements for AI infrastructure. The focus on training efficiency alone doesn’t fully account for the broader technology ecosystem.
Finally, the integration of DeepSeek’s R1 model into Nvidia’s platform demonstrates the company’s ability to adapt to market developments while maintaining its premium positioning in the AI chip sector.
Stay Ahead of the Market with Benzinga Pro!
Want to trade like a pro? Benzinga Pro gives you the edge you need in today's fast-paced markets. Get real-time news, exclusive insights, and powerful tools trusted by professional traders:
- Breaking market-moving stories before they hit mainstream media
- Live audio squawk for hands-free market updates
- Advanced stock scanner to spot promising trades
- Expert trade ideas and on-demand support