The emergence of Chinese artificial intelligence startup DeepSeek has sent shockwaves through global technology markets, triggering a significant selloff in AI-driven stocks and raising questions about the long-standing dominance of U.S. technology companies.
The startup’s groundbreaking AI model, which runs cost-effectively on less-advanced chips, has cast doubt on the sustainability of high valuations in the sector and the strategies of industry leaders like Nvidia Corp.
DeepSeek unveiled a groundbreaking reasoning model, R1, that operates cost-efficiently on less-advanced chips. The release of R1, developed for less than $6 million using Nvidia’s H800 chips, bypasses U.S. export restrictions that bar China from acquiring more powerful H100 chips, marking a significant shift in the AI industry.
Register for Tekedia Mini-MBA edition 17 (June 9 – Sept 6, 2025) today for early bird discounts. Do annual for access to Blucera.com.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register to become a better CEO or Director with Tekedia CEO & Director Program.
This breakthrough sent shockwaves through global markets, with Nvidia Corp. (NVDA) leading a steep selloff as its shares plummeted more than 17% in trading on Monday. Futures on the Nasdaq 100 fell 3.4%, while S&P 500 contracts dropped 2% in premarket trading.
The ripple effect extended to Europe, where tech stocks led market losses, with ASML Holding NV, a critical supplier of semiconductor manufacturing equipment, tumbling 11%. Combined, the Nasdaq 100 and Europe’s Stoxx 600 technology sub-index faced a potential $1 trillion market capitalization wipeout if losses hold.
DeepSeek’s R1 model demonstrates the ability to achieve high performance with significantly lower capital and hardware requirements, challenging the premise that cutting-edge AI models necessitate advanced, expensive chips like Nvidia’s H100. By leveraging the less-powerful H800 chips—designed to comply with U.S. export restrictions—DeepSeek showcased an alternative pathway to AI innovation.
“This breakthrough redefines what’s possible in AI,” said Vey-Sern Ling, managing director at Union Bancaire Privee. “DeepSeek shows that it is possible to develop powerful AI models that cost less. It can potentially derail the investment case for the entire AI supply chain, which is driven by high spending from a small handful of hyperscalers.”
R1’s development cost of under $6 million starkly contrasts the billions poured into AI research by Silicon Valley companies, casting doubt on the sustainability of their high-spending models.
Investor jitters were evident as trading volumes surged. By 4:45 a.m. New York time, roughly 200,000 Nasdaq 100 futures contracts had changed hands—four times the 30-day average for this time of day. The Cboe Volatility Index (VIX), often referred to as Wall Street’s “fear gauge,” spiked higher, reflecting heightened uncertainty.
“This is deeply problematic for the thesis that significant capital expenditure and operating expenses are the best ways to approach the AI trend,” said Nirgunan Tiruchelvam, head of consumer and internet at Aletheia Capital.
The selloff coincides with a crucial week for tech earnings, including reports from Apple Inc. and Microsoft Corp. Analysts predict slowing profit growth amid inflated valuations, compounding concerns about the sector’s resilience.
The Nasdaq 100, currently trading at 27 times estimated forward earnings compared to a three-year average of 24, appears particularly vulnerable. Nvidia’s valuation, at 33 times estimated forward earnings, underscores the risk of overreliance on a single narrative: that AI innovation is inextricably tied to high-cost infrastructure.
China’s Rise in AI
DeepSeek’s success challenges the prevailing notion that China’s AI sector lags significantly behind its U.S. counterparts. It is especially significant as it bypasses Washington’s restrictions on exporting cutting-edge chips to China.
“While current leaders like Nvidia have a strong foothold, it is a reminder that AI dominance cannot be taken for granted,” said Charu Chanana, chief investment strategist at Saxo Markets. “The emergence of China’s DeepSeek indicates that competition is intensifying, and future competitors will evolve faster and challenge established companies more quickly.”
Chinese AI-related stocks reacted positively. Mainland-listed companies with links to DeepSeek, such as Merit Interactive Co., surged by their daily trading limits. In Hong Kong, the Hang Seng Tech Index climbed 2%, buoyed by optimism about DeepSeek’s potential.
A Game-Changing Model
R1’s standout feature is its transparency. Unlike many existing AI models, which operate as black boxes, R1 demonstrates its reasoning process, providing users with a clearer understanding of its conclusions. This feature propelled DeepSeek’s app to the top of Apple Inc.’s App Store rankings within days of its release, garnering praise from users and investors alike.
Notable investor Marc Andreessen described R1 as “one of the most amazing and impressive breakthroughs,” highlighting its potential to reshape the AI landscape.
For years, Silicon Valley has championed the idea that advanced AI requires substantial investments in cutting-edge hardware and energy-intensive infrastructure. DeepSeek’s R1 upends this narrative, showing that cost-effective solutions can rival, and perhaps, surpass the capabilities of resource-heavy alternatives.
“This development calls into question the massive resources dedicated to AI by Silicon Valley,” Tiruchelvam added. “It forces a reassessment of whether these expenditures are justified.”
Tech analysts note that the success of DeepSeek underscores the growing competitiveness of China’s AI sector and the potential for open-source innovation to bridge technological gaps. It also highlights the vulnerabilities of U.S. tech companies, which have built their business models around the assumption of sustained dominance in AI hardware.
“The AI supply chain is being fundamentally disrupted,” Ling said. “This isn’t just about DeepSeek; it’s about the future of AI development and the assumptions that underpin it.”
NVIDIA in a statement on Monday says DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling.
“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. Inference requires significant numbers of NVIDIA GPUs and high-performance networking. We now have three scaling laws: pre-training and post-training, which continue, and new test-time scaling,” it said.
NVIDIA lost $580 billion in market value on Monday.