Nvidia launches new generation "super chip" for artificial intelligence

Nvidia CEO Jensen Huang opened the company's 2024 Annual Conference in California on March 18 with notable announcements aimed at securing the chipmaker's dominant position in the field of artificial intelligence (AI).

Speaking on stage in the heart of Silicon Valley, Jensen Huang announced Nvidia's latest generation of chips called "GB200" with the ability to perform tasks 30 times faster than the previous generation (H100) in certain tasks, while consuming up to 25 times less power.

According to Nvidia, the new generation of AI graphics processors based on the Blackwell architecture can bring a huge performance upgrade for businesses applying AI.

Specifically, the computing capacity of GB200 is 20 petaflops, 5 times larger than the 4 petaflops of H100 - the chip generation that is always "out of stock" in 2023.

Nvidia claims this additional processing power will allow AI companies to train larger and more complex models.

Nvidia said its major customers Amazon, Google, Microsoft, OpenAI and Oracle are expected to use the new generation of chips in cloud computing and AI services.

CEO Jensen Huang's announcements will help determine whether Nvidia can maintain its 80% share of the AI ​​chip market.

Nvidia has become a $2 trillion company by providing the crucial chips used to train complex AI models.

The H100 is the most powerful processor Nvidia has ever built based on the Hopper architecture, and is one of the most highly regarded chip generations in the technology world, priced at $30,000-40,000 per chip.

This American "semiconductor giant" is also preparing to join hands with Japanese electronics corporation Hitachi to develop a new AI system.

According to a VNA reporter in Tokyo, the first phase of cooperation between the two "technology giants" will take place from April to June with the sale of AI-equipped data storage servers.

The system is equipped with Nvidia's image processing unit (GPU) so it can process large amounts of complex data generated by AI.

Hitachi said it will work with Nvidia to develop a “digital twin” system to replicate infrastructure and manufacturing sites in the metaverse.

By combining Hitachi's equipment operation and control know-how with Nvidia's virtual space and AI technology, the system enables simulation of equipment and machinery installation in virtual space.

Hitachi aims to expand its digital transformation (DX) support business as a pillar of growth going forward.