The CEO of Nvidia, Jensen Huang, made this announcement at the Taipei Music Center, site of the Calpex AI exposure (file) | Photo credit: AP
NVIDIA said on Monday that it planned to sell a technology to others that will bind fleas to accelerate communication from chip to chip necessary to build and deploy artificial intelligence tools.
NVIDIA launched a new version of its NVLink Tech called NVLink Fusion Monday that it will sell to other flea designers to help create powerful personalized AI systems with several chips linked together.
Marvell Technology and Mediatek plan to adopt NVLink technology called Fusion for their personalized flea efforts.
NVIDIA has developed NVLink years ago, and it is used to exchange massive amounts of data between various chips, as in the company’s GB200, which combines two Blackwell graphic treatment units with a Grace processor.
The CEO of Nvidia, Jensen Huang, made this announcement at the Taipei Music Center, site of the computex exhibition aiwhich takes place from May 20 to 23.
In addition to the new technological production, Huang announced the plan of the company to build a head office in Taiwan in the northern suburbs of Taipei.
His opening speech discussed Nvidia’s story to build ia fleas and systems and the software that the company has built to support it.
Huang pointed out that at one point, his presentations spent 90% of their time on the company’s graphic fleas, but that has changed.
Now, Nvidia has increased beyond its roots as a manufacturer of video game graphics in the dominant producer of fleas who have propelled the frenzy of the AI which has entered the technological industry since the launch of Chatgpt in 2022. Nvidia has designed processors that previously published the Windows operating system from Microsoft and use Arm Holdings technology, previously announcement.
In Computex last year, Huang sparked “Jensanity” in Taiwan, while the public and the media followed the CEO, which was assailed by the participants in the show. During the annual conference of the company’s developers in March, Huang explained how Nvidia would position itself to meet the change in computer needs, from the construction of large AI models to the execution of applications according to them.
He has unveiled several new generations of AI chips, including the Blackwell Ultra, which will be available later this year.
The Rubin fleas of the company will be followed by Feynman processors, which should arrive in 2028.
Nvidia also launched an office version of its AI chips, called DGX Spark, targeting AI researchers. Huang said on Monday that the computer was in full production and would be ready in a few “a few weeks”.
Calpex, who should have 1,400 exhibitors, will be the first major gathering of computer leaders and chips in Asia, because the American president Donald Trump threatened to impose radical prices to push companies to increase production in the United States
Published – May 19, 2025 11:22 AM Ist