How DeepSeek’s AI disruption sent a $593 billion shockwave to Nvidia, rattled global stock markets

How DeepSeek’s AI disruption sent a $593 billion shockwave to Nvidia, rattled global stock markets

The launch of DeepSeek's cost-effective AI assistant in China has sent shockwaves through global markets, leading to a record $593 billion loss for Nvidia and a 3.1 per cent drop in the Nasdaq. As investors recalibrate, the implications for the tech industry and AI landscape are profound.

DeepSeek AI , a low-cost Chinese AI start-up from Hangzhou, caused a major shake-up in global markets, wiping $593 billion off Nvidia’s value in a single day a record 17% drop. This is the largest one-day loss for any Wall Street stock, according to LSEG data.

The disruption hit Nvidia, led by Jensen Huang, the hardest, but also dragged down the Nasdaq by over 3% and affected the S&P 500. Investors panicked over DeepSeek's low-cost, open-source AI rivalling costly U.S. AI efforts, shaking confidence in the tech industry worldwide.

What is DeepSeek AI?

Chinese start-up DeepSeek made waves by launching a free AI assistant last week, claiming it operates with less data and at a fraction of the cost compared to rivals like OpenAI’s ChatGPT and Google’s Gemini.

By January 27, DeepSeek AI had overtaken ChatGPT in downloads on the US iOS App Store and surpassed it in key regions, including the UK, Australia, Canada, China, and Singapore, according to Reuters.

Why Is China's DeepSeek a Threat to US AI Players?

DeepSeek, a Hangzhou-based start-up majority owned by Liang Wenfeng, co-founder of hedge fund High-Flyer, has disrupted the AI landscape with its models DeepSeek-V3 and DeepSeek-R1.

Last month, benchmarks revealed that the V3 large language model outperformed those of major US tech giants, despite being developed at a much lower cost. Meanwhile, the R1 model, designed to mimic human reasoning, matched and even outperformed OpenAI’s latest o1 model in several tests, solidifying DeepSeek’s rapid rise to prominence.

In a recent paper, DeepSeek revealed that its V3 model was trained using Nvidia H800 chips at a cost of under $6 million far less than the billions spent by AI giants like Microsoft, Meta, and OpenAI.

The company’s success has also sparked debate over the effectiveness of US export restrictions on advanced AI chips. With its low-cost, high-performance models, DeepSeek threatens to cut into the market share of major AI players while potentially driving down prices across the industry.

How Have Tech Stocks Reacted? Nvidia, Broadcom, Microsoft and Google Tumble

On January 27, the Nasdaq dropped 3.1%, led by Nvidia's record 17% loss, sparked by DeepSeek's rise. Chipmaker Broadcom fell 17.4%, Microsoft dropped 2.1%, and Alphabet lost 4.2%, according to Reuters. The Philadelphia semiconductor index plunged 9.2%, its biggest drop since March 2020. Marvell Technology suffered the steepest decline on the Nasdaq, tumbling 19.1%.

The global selloff on January 27 began in Asia, with Japan's SoftBank Group dropping 8.3%, and spread to Europe, where ASML fell 7%. In the US, Vertiv Holdings, a data center infrastructure company, plunged 29.9%. Power companies, including Vistra (-28.3%), Constellation Energy (-20.8%), and NRG Energy (-13.2%), also suffered as investors reversed bets on a surge in demand from AI-driven data centers.

Expert View: Is DeepSeek A Threat to US Tech Cos?

Jefferies has highlighted that DeepSeek’s open-source LLM, which matches GPT-4o’s performance using far less computing power, could shift the AI industry's focus to return on investment (ROI). Their report, "The Fear Created by China's Deepseek", noted that with a training cost of $5.6 million, DeepSeek’s V3 is 10% cheaper than Meta’s Llama and outperforms Llama 3.1 while matching GPT-4o. The model, freely available on Hugging Face, allows AI developers to create applications at much lower costs. However, DeepSeek has not yet focused on commercializing its technology.

Brian Jacobsen, an economist at Annex Wealth Management, stated that if DeepSeek's claims hold, it could be the “better mousetrap” that disrupts the AI narrative, reducing demand for chips, power, and large-scale data centers. Former US President Donald Trump weighed in, calling DeepSeek a potential “wakeup call” and a positive development for the industry.

What did US stock market investors buy on Jan 27?

According to Reuters, most of the outflows from tech stocks shifted into safer investments like government bonds and currencies. The US Treasury 10-year yield dropped to 4.53%, while Japan’s Yen and the Swiss Franc strengthened against the US Dollar. A trading executive noted that the increased volatility in tech stocks could prompt banks to adjust their risk management strategies, potentially holding fewer tech shares or managing positions more cautiously as clients unwind their holdings.

What Did rival AI ChatGPT's Sam Altman say about DeepSeek?

OpenAI founder Sam Altman responded positively to DeepSeek’s rise, calling it “invigorating” to have new competition. In a post on X, Altman acknowledged DeepSeek’s R1 model as impressive, especially given its price, and hinted that OpenAI would release even better models to compete. He also emphasized the importance of continued research and development, saying, “More computing is more important now than ever before” and expressed excitement about the next-generation AI models that will amaze the world.

#DeepSeekAI #AIcompetition #NvidiaCrash #TechStocks #OpenAI #GPT4 #AIrevolution #ArtificialIntelligence #TechDisruption #AIindustry #DeepLearning #AIbenchmark #AIstartup #AIcosts #SamAltman #JensenHuang #AIinvesting #TechSelloff #NasdaqDrop #USTechStocks #AIinnovation #Tradesala #SupportIndianBrands #SupportIndianProducts

要查看或添加评论,请登录

Tradesala - Shop Indian Brands的更多文章

社区洞察

其他会员也浏览了