The much-anticipated “Blackwell” B-200 AI chip from Nvidia is going to be late, which is creating uncertainty in the AI business. Two other people and a Microsoft employee told The Information that the chip’s release has been pushed back by at least three months because of a design flaw.
The CEO of Nvidia, Jensen Huang, recently said that engineering samples would be sent out starting July 31 at the SIGGRAPH event in Denver, Colorado. However, the company had not publicly announced a launch date.
Still not clear is whether the delay will also affect the B-100 class along with the B-200. The architecture of both series is similar, but the B-200 is expected to work better.
Nvidia Faces Demand Pressure
Analysts think that Nvidia will sell enormous quantities of B-series AI chips to significant names like Amazon, Google, Meta, Microsoft, and other AI companies. The AI industry as a whole and Nvidia’s 2025 earnings will likely rest on how well they can meet this demand.
With a market capitalization of $2.6 trillion, Nvidia is by far the market leader. Its closest competitor in the semiconductor business, TSMC, has a market cap of $777 billion. TSMC, on the other hand, is more of a strategic partner and makes most of Nvidia’s chips.
Nvidia does not face much competition in the U.S. Microsoft and Google are working on their own AI chips, but Nvidia is still the best in the field. Intel and AMD, two competitors, have had a hard time getting a position in the generative AI market and are now shifting their focus to the AI market.
Major tech has had a rough second quarter of 2024. As of August 3, 2024, the market value of all but two of the top 20 global tech companies had gone down. Nvidia is following this trend, even though its price hit a record high in 2024.
After Nvidia’s rise since OpenAI’s ChatGPT launch, which used Nvidia chips, investment experts are anxious for the next significant thing to happen. According to the Financial Times, Elliott Management, a hedge fund, told investors that Nvidia was “in bubble territory” and that they were skeptical about the cost-effectiveness and dependability of current uses of generative AI.
Sequoia Capital, a venture capital firm, also gave the AI industry an adverse outlook, saying that AI companies would need to make about $600 billion in sales to support the funds they spend on Nvidia GPUs.
