DeepSeek storm sweeps through US tech giants, Wedbush: Don't panic, it's a golden buying opportunity!

Zhitong
2025.01.27 14:02
portai
I'm PortAI, I can summarize articles.

DeepSeek's R1 large language model has triggered a sell-off in U.S. tech stocks, with Wedbush Securities considering this a "golden buying opportunity" for investors. Analysts pointed out that despite the impact of the DeepSeek model on the market, U.S. tech companies' dominance in the field of general artificial intelligence remains unshakable. Stocks of tech companies such as NVIDIA and Microsoft have all seen declines, while DeepSeek's AI assistant has become the top free app on the App Store

According to the Zhitong Finance APP, Wedbush stated that on Monday, the sell-off of technology stocks triggered by DeepSeek's R1 large language model and market concerns about the cost of competitive models for U.S. companies may provide investors with a "golden buying opportunity." Wedbush analysts commented in a report, "Clearly, technology stocks are facing immense pressure led by NVIDIA (NVDA.US), and Wall Street will view DeepSeek as a significant threat to U.S. technological dominance and control over this artificial intelligence revolution."

As of the time of writing, NVIDIA's stock price fell 11% in pre-market trading, while Microsoft (MSFT.US), which holds shares in generative AI startup OpenAI, dropped 4%. Other AI-related stocks were also hit hard, including Broadcom (AVGO.US) down 11%, and AMD (AMD.US) down 4%. Meta Platforms (META.US) and Google (GOOGL.US) fell 3% and 3.5%, respectively.

Wedbush analysts said, "While the DeepSeek model is impressive and has a ripple effect on the market, the reality is that the seven giants of the U.S. stock market and U.S. tech companies are focused on the final stages of (general artificial intelligence) and all the infrastructure and ecosystems that are beyond the reach of China and DeepSeek. Currently, the focus of AI is on enterprise use cases and broader infrastructure, which will drive $2 trillion in capital expenditures over the next three years."

DeepSeek released its free open-source large language model R1 at the end of December, claiming it was developed in two months at a cost of $5.57 million. DeepSeek stated that it used NVIDIA's H800 GPU for approximately 2,788 hours. Currently, DeepSeek's AI assistant has replaced ChatGPT as the number one free app on the App Store.

Meanwhile, some individuals have expressed skepticism, stating that research shows the aforementioned costs only include the "official training" of DeepSeek-V3, excluding costs related to prior research and ablation experiments concerning architecture, algorithms, and data.

Wedbush Securities stated that it remains to be seen whether DeepSeek can find a way to circumvent U.S. export controls on NVIDIA and other GPU manufacturers. The agency added that while the launch of the DeepSeek model is impressive, ultimately, it is not seen as a "competitive threat" necessary to continue building a broader AI infrastructure

Source: Zhitong The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.