Nvidia's new AI chip delayed, asset management company: Company in bubble stage, AI overhyped

The Information reported that due to design flaws, the release date of Nvidia's new "Blackwell" AI chip has been delayed by at least three months, affecting companies including Microsoft and Google. In response, asset management companies believe that the company is currently in a bubble, as AI has been excessively hyped.

Nvidia's New AI Chip Delayed Due to Flaw

Reports citing an anonymous Microsoft employee and two other sources indicate that Nvidia has notified Microsoft and other cloud service providers that its "Blackwell" B200 AI chip has been found to have flaws during production, causing a delay of at least three months:

Nvidia is currently conducting a series of new tests with another chip manufacturer, TSMC, and expects to start shipping the Blackwell chips in bulk in the first quarter of next year.

It is also noted that "production volume and speed will gradually increase and accelerate in the latter part of this year."

This delay will primarily affect tech companies like Microsoft, Google, and Meta, who have reportedly ordered billions of dollars worth of chips from Nvidia, as claimed.

It is reported that the B200 chip is the next-generation flagship product following the H100 chip, which sparked an unprecedented AI wave, becoming a major driving force in the fields of AI, cloud, and computing services, and securing an impressive market share for Nvidia's chip manufacturing giants. This is why the B200 is highly anticipated.

AMD Exceeds Expectations in Financial Report, Targeting Nvidia

Following the release of Q2 financial reports last week, most tech companies, apart from Apple, among the top 20 global tech companies by market value, faced significant declines.

However, AMD, one of Nvidia's main competitors, showcased a strong financial report at this time.

From the launch of the new AI flagship product, the MI 300 accelerator chip, to the acquisition of AI startup Silo AI, it is evident that the company is challenging Nvidia's dominant position in the AI field.

Su Zifeng announces AMD's profit report: MI 300 chip poses a strong challenge to Nvidia, AI deployment

A few days ago, Apple adopted Google's cloud chip TPU as the basis for the company's AI model training, indirectly verifying the idea that tech companies involved in the AI field are seeking alternatives to Nvidia chips.

Is Apple Breaking Up with Nvidia? New AI System Reportedly Collaborates with Google

Asset Management Company: Nvidia in Bubble, AI Overhyped

Furthermore, asset management company Elliott, managing over $70 billion in assets, believes that Nvidia is currently in a bubble due to the overhype of AI technology.

The company emphasized in a letter to clients that AI has been overhyped and has yet to provide software or products of corresponding value:

Many use cases for AI "will never have significant benefits, will never function properly, will only consume excessive energy, or will be proven untrustworthy."

It also added, "Many AI applications are not yet ready, with high costs and instability."

Elliott criticized that aside from "analyzing meeting records, generating reports, and assisting in computer coding," AI has no other practical uses,

As for when the market bubble may burst, the company stated:

If Nvidia announces poor performance or data in a report, this situation could occur.