DeepSeek: What is DeepSeek? China’s AI Challenger to OpenAI and Nvidia

SPV
SPV
5 Min Read
hostinger buy

DeepSeek: What is DeepSeek? China’s AI Challenger to OpenAI and Nvidia

DeepSeek is a private Chinese company founded in July 2023 by Liang Wenfeng, a Zhejiang University graduate. Liang funded the startup through his hedge fund, with an estimated $8 billion in assets.

Initially focused on applying AI to investing, Liang acquired a significant stockpile of Nvidia A100 chips, which have since been banned for export to China. These chips became the foundation of DeepSeek’s AI technology.

According to a Jan. 27 research note by Ives, DeepSeek has made significant strides in AI breakthroughs, particularly in areas like improved learning and more efficient memory use. However, some skepticism remains about the company’s reliance on a large number of chips.

Is DeepSeek Available in the U.S.?

Yes, DeepSeek’s AI app is available for download in the Apple App Store and can also be accessed online via its website. As of January 28, it was the top download on the App Store, although some users faced difficulties during registration. The Chinese version of the service faced “large-scale malicious attacks,” which prompted DeepSeek to temporarily limit new sign-ups. Existing users can still log in.

hostinger buy

DeepSeek released a new AI model on January 20, which is sparking a reevaluation of the AI sector. The model, which competes with OpenAI’s ChatGPT and Meta’s Llama 3.1, has caught Wall Street’s attention. According to Wedbush’s Ives, DeepSeek’s use of reduced capability Nvidia chips is a noteworthy achievement, which has contributed to pressure on U.S. tech stocks.

How is DeepSeek Different from Other AI Apps?

DeepSeek stands out due to its use of inference-time computing. This approach activates only the most relevant parts of its AI model for each query, making the process more efficient in terms of both cost and computational power.

Experts have praised DeepSeek’s performance. Tech investor Marc Andreessen called DeepSeek R1 a breakthrough, stating it is “one of the most amazing and impressive breakthroughs I’ve ever seen.” He also praised DeepSeek’s decision to make the model open-source, calling it “a profound gift to the world.”

David Sacks, a venture capitalist, also commented on DeepSeek’s potential, saying that it demonstrates how competitive the AI race has become. Despite this, Ives expressed doubts that DeepSeek will gain significant traction among major U.S. companies. He argued that U.S. businesses would be unlikely to adopt a Chinese startup’s AI tools for large-scale infrastructure.

What Does DeepSeek Mean for Nvidia and Other Tech Companies?

DeepSeek’s AI advancements are raising concerns among U.S. tech companies, especially Nvidia. The fact that DeepSeek’s technology is more cost-effective and energy-efficient than models from U.S.-based companies has sparked worry on Wall Street.

Some analysts believe that U.S. companies are overspending on AI technology, while others think the long-term demand for AI infrastructure will continue to benefit key players like Nvidia. Jay Woods, Chief Global Strategist at Freedom Capital Markets, pointed out that while DeepSeek’s model is cheaper to run, the vast infrastructure needed to support the growing demand for AI still favors companies with significant resources.

This situation is causing analysts to reconsider whether the enormous AI investments being made in Silicon Valley will continue to be justified.

Nvidia’s Reaction to DeepSeek

Nvidia acknowledged DeepSeek’s AI advancements, calling it a great example of “test-time scaling.” The company emphasized that DeepSeek’s approach demonstrates how new models can be developed by leveraging available models and compute resources in a way that complies with export controls.

However, Nvidia clarified that true AI inference—making predictions or decisions based on AI models—requires substantial amounts of Nvidia GPUs and high-performance networking infrastructure. The company also highlighted the three “scaling laws” of AI: pre-training, post-training, and test-time scaling, which underscore the ongoing need for Nvidia’s advanced hardware.

In summary, DeepSeek is an emerging Chinese AI company that is shaking up the tech landscape with a model that rivals established players like OpenAI and Meta. While the company has gained attention for its cost-effective and efficient AI tools, it remains to be seen how it will impact major U.S. tech companies and Nvidia in particular.

hostinger buy
Share This Article
Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *