AI: Nvidia launches supercomputer and resource for games – 05/29/2023 – Tech

Nvidia Chief Executive Jensen Huang announced a suite of the company’s artificial intelligence (AI) products on Monday. New features include AI capabilities for cars, robots, industry, and realistic character generation for video games.

Huang also announced the launch of Nvidia’s AI supercomputer, DGX GH200, with 256 superchips equipped with the H100 graphics processing unit – sold at R$ 200,000 per unit. The company’s previous model, DGX AI, had 500 times less memory, according to the chip maker.

The executive opened the Computex technology forum in Taiwan on Monday when he made the announcements. The day before, he spoke to graduates at the National Taiwan University in Taipei.

“While some are worried that artificial intelligence will take away their jobs, someone who is an expert in AI will do it,” Huang said. Those who do not embrace technology will be left behind.

The event served as the stage for Huang’s announcement of the new line of infrastructure aimed at artificial intelligence. Nvidia launched, in addition to supercomputers, resources for cloud computing and internet acceleration. Among the company’s clients are Google, Microsoft, Meta, Dell and Lenovo.

The executive even brought more tangible applications to the public. The Nvidia Avatar Cloud Engine (ACE) for Games will assist in the creation of video game characters, with appearance, animation and voice. The product aims to serve developers and game studios.

Nvidia has also partnered with multimedia company Mediatek to bring artificial intelligence capabilities to connected cars. Huang says this technology should be available from entry-level cars to luxury vehicles.

The package also featured news about the Nvidia Omniverse —the company’s metaverse dedicated to the corporate world. This mixed reality space allows simulations and inspections in the work environment.

The generative artificial intelligence will make it possible to develop programs on the Nvidia Omniverse without using code, only with commands in virtual language.

In addition, Nvidia announced the Isaac MVR mobile robot, which locates and moves using information from the Omniverse. Along the same lines, robotic cameras will be able to inspect equipment autonomously.

Huang did not disclose prices during the technology event in Taipei, but he said that 40,000 large companies and 15,000 startups use Nvidia technology.

Nvidia’s processing software, Cuda, was downloaded 25 million times just last year. The company’s success is reflected in the balance sheet for the first quarter of 2023, with revenue of US$ 7.19 billion, according to Nvidia’s director for Latin America, Marcio Aguiar.

The company’s results made Nvidia’s shares appreciate by 25% last Thursday (25). With that, the company flirts with a market value of US$ 1 trillion (R$ 5 trillion), reached by technology giants, such as Apple and Microsoft, and the Saudi oil company Aramco.

Previously identified with chips for games, the American Nvidia has become a bet on Wall Street to invest in artificial intelligence, because it also produces GPUs (Graphics Processing Units) that are used to train language models for generative AI, such as ChatGPT.