Nvidia's Stock Rally Pauses. A New Generation of Data Centers Is Arriving

Dow Jones09-16

Nvidia was slipping early on Monday as the chip maker looks set to give back some gains from the previous week's rally.

Nvidia shares were down 1.44% at $117.39 in premarket trading on Monday. The stock closed broadly flat on Friday but has risen 12% over the past five trading sessions.

Nvidia has gained from renewed confidence in demand for its chips, partly fueled by cloud-computing and software company Oracle which has talked of companies spending $100 billion each on developing their artificial-intelligence models.

Oracle has outlined plans to build an AI supercomputer that will be powered by 131,072 of Nvidia's next-generation Blackwell chips

"Nvidia is at the epicenter of all things related and is the pioneer for artificial intelligence infrastructure," wrote Ken Mahoney, CEO of Mahoney Asset Management, in an emailed comment. "They have been and will continue to be the beneficiary of this movement."

However, there are signs that investors in big data centers are diversifying what type of chips they intend to use. Chip start-up Groq said recently it has joined with oil producer Saudi Arabian Oil Co., commonly known as Aramco, to build a giant data center in Saudi Arabia to offer AI computing power to local companies.

Groq's chips are focused on inference, the process of generating answers or results from AI models. It has claimed its language-processing units are better suited for powering AI language applications than GPUs of the type Nvidia makes.

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

  • Icywinddale
    09-16
    Icywinddale
    The writer knows nothing
  • kibkibkib
    09-16
    kibkibkib
    Inference chips are different from the GPU that supports CUDA for all AI applications not limited to inference alone, and no technical comparison between GPU and their inference chips performance. Most of the money spent by big tech companies are spent on training models not inference.
Leave a comment
2
1