ECA
01-26 16:23

Dude. What are you talking about man. NVDA and DeepSeek are not competitors and produce different products and services. DeepSeek threatens the LLMs produced by other hyper scalers like Google Gemini, OpenAi, Llama, etc. If anything you admitted that it actually runs on NVDA chips, and so it needs NVDA. DeepSeek actually strengthen's NVDA's dominance. 

Will Chinese DeepSeek Disrupt Nvidia?
DeepSeek Forces Nvidia Down? AI Logic Changing?
On January 24th, a thread posted on the anonymity platform teamblind went viral. A Meta employee claimed that Meta is now in panic mode internally because of DeepSeek's models. Some investors have even pointed to DeepSeek's blowout as the reason for Nvidia's fall on Friday: because the company appears to have built a groundbreaking AI model at a fraction of the cost and without the use of cutting-edge chips. Do you think DeepSeek's emergence will be a short-term or long-term bearish factor for Nvidia? Would you buy the dip or wait?
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

  • ECA
    01-26 17:51
    ECA
    And DeepSeek is like what one of many NVDA’s clients? It doesn’t even serve the Chinese market that much. It’s not even one of the biggest customers. NVDA can service the whole domestic market in US and doesn’t really need the rest of the world to still grow its market. Data Centre revenue is massive. Does DeepSeek produce chips like NVDA does? It’s not a competitor dude. In the future, of course compute will become more efficient but then so will the cost of chips. It’s inevitable.
  • SmoothTW
    01-26 17:20
    SmoothTW
    The new approach to LLM training that Deepseek R1 used is much less compute intensive than other models. Less compute = less need for capex on high end GPUs = bearish for NVDA in the short term.
Leave a comment
2