Nvidia presented to Taipei Computex, one of the world’s leading conferences on computer innovations, a new supercomputer based on artificial intelligence (AI). It’s about the Dgx Gh200, which will allow large companies to accelerate the development and distribution of their AI projects.
Nvidia and the supercomputer for AI presented in Taipei
The new supercomputer will make it possible to develop AI solutions, accessible to many customers. Among Nvidia’s customers, as announced by the CEO Jensen HuangThere are Google Cloud, Meta e Microsoft.
With the supercomputer they will be able to create applications with advanced chatbots, tools to monitor internet security and realistic video games with characters that adapt to the context and interact with the players.
The template was used by Nvidia to create Helios, an early operational supercomputer. This is formed by four Dgx Gh200 systemseach of which can handle large AI workloads.
A single system contains 256 superchipcalled Grace Hopperto create a unique machine, capable of doing any AI job.
Grace superchips are the main component of Helios’s ‘brother’ supercomputer, Taiwania 4, built by Asus for the Taiwan National Center for High-Performance Computing. A sign that the great computational needs of artificial intelligence are pushing manufacturers to focus on supercomputers.