Nvidia unveils latest chips, technology to speed up AI computing

Nvidia’s new Grace CPU chip unveiled at the chipmaker’s AI Developer Conference can be seen in this undated image obtained by Reuters. Nvidia/Handout via REUTERS

Join now for FREE unlimited access to Reuters.com

Register

March 22 (Reuters) – Nvidia Corp on Tuesday announced several new chips and technologies that it says will increase the computing speed of increasingly complicated artificial intelligence algorithms, intensifying competition against manufacturers chip rivals vying for lucrative data center businesses.

Nvidia’s graphics chips (GPUs), which initially helped propel and improve video quality in the gaming market, have become the dominant chips companies use for AI workloads. The latest GPU, called the H100, can help cut compute times from weeks to days for certain jobs that involve training AI models, the company said.

The announcements were made during Nvidia’s online AI Developer Conference.

Join now for FREE unlimited access to Reuters.com

Register

“Data centers are becoming AI factories – processing and refining mountains of data to produce intelligence,” Nvidia CEO Jensen Huang said in a statement, calling the H100 chip “the engine” of AI infrastructure.

Businesses are using AI and machine learning for everything from recommending which video to watch next to discovering new drugs, and the technology is increasingly becoming an important tool for business.

The H100 chip will be produced on Taiwan Manufacturing Semiconductor Company’s state-of-the-art four-nanometer process (2330.TW) with 80 billion transistors and will be available in the third quarter, Nvidia said.

The H100 will also be used to build Nvidia’s new “Eos” supercomputer, which Nvidia says will be the world’s fastest AI system when it goes live later this year.

Facebook parent company Meta (FB.O) announced in January that it would build the world’s fastest AI supercomputer this year and that it would run at nearly 5 exaflops. Nvidia said Tuesday that its supercomputer will operate at more than 18 exaflops.

Exaflop’s performance is the ability to perform 1 quintillion – or 1,000,000,000,000,000,000 – calculations per second.

Nvidia has also introduced a new processor (CPU) chip called Grace CPU Superchip, which is based on Arm technology. It’s the first new chip from Nvidia that uses the Arm architecture to be announced since the company’s deal to buy Arm Ltd fell apart last month due to regulatory hurdles.

The Grace CPU Superchip, which will be available in the first half of next year, connects two CPU chips and will focus on AI and other compute-intensive tasks.

More and more companies are connecting chips using technology that enables faster data flow between them. Earlier this month Apple Inc unveiled its M1 Ultra chip connecting two M1 Max chips. [nL2N2VB1DI]

Nvidia said the two CPU chips are connected using its NVLink-C2C technology, which was also unveiled on Tuesday.

Nvidia, which developed its self-driving technology and expanded the business, said it began shipping its “Drive Orin” self-driving vehicle computer this month and Chinese electric vehicle maker BYD Co Ltd (002594.SZ) and luxury electric car maker Lucid Motors is reportedly using Nvidia Drive for its next-gen fleets.

Danny Shapiro, Nvidia’s vice president for automotive, said there was $11 billion of automotive business in the “pipeline” over the next six years, up from the $8 billion he expected last year. ‘last year. The anticipated revenue growth will come from Nvidia’s hardware and increased recurring software revenue, Shapiro said.

Nvidia shares were relatively flat in midday trading.

Join now for FREE unlimited access to Reuters.com

Register

Reporting by Jane Lanhee Lee, additional reporting by Joseph White; Editing by Bernard Orr

Our standards: The Thomson Reuters Trust Principles.

About Dianne Stinson

Check Also

i-Synergy creates new technologies to keep up with megatrends in digital e-commerce and marketplaces

VTRAK and Rewalty’s i-Synergy integrated platform and hybrid marketplace will capture the ongoing transformation in …