US chip giant Nvidia unveils new AI chip

Nvidia CEO Jensen Huang, in his speech at the company's GPU Technology Conference (GTC), explained that the amount of data available will increase and that even larger graphics processing units (GPUs) will need to be created.

Stating that the company's current chip architecture, Hopper, is "fantastic" but that they need larger GPUs, Huang introduced the new generation chip architecture, which takes its name from mathematician David Blackwell.

Huang noted that Blackwell is not a chip, but the name of a platform.

On the other hand, the company's statement reported that the Blackwell platform allows for the creation and operation of real-time generative AI on trillion-parameter large language models with up to 25 times less cost and energy consumption compared to the previous model.

The statement noted that companies expected to use Blackwell include Amazon Web Services, Dell Technologies, Google, Meta, Microsoft, OpenAI, Oracle, Tesla, and xAI.

Furthermore, it was reported that the first Blackwell chip, the GB200, will be released later this year.

News Source: AA