(Reuters)-Nvidia said on Tuesday it will release a new artificial intelligence chip by the end of next year, designed to deal with complex functions such as video and software creation.
The chips, named “Rubin CPX”, will be built based on Nvidia’s next -generation Rubin architecture – the successor of its latest Blackwell technology, which marked the company’s incursion in the supply of larger processing systems.
As IA systems become more sophisticated, dealing with heavy tasks such as “vibration coding” or AI -generating video code and video generation, industry processing needs are intensifying.
Free tool
XP Simulator
Learn in 1 minute how much your money can yield
AI models can take up to 1 million tokens to process an hour of video content – a challenging feat for traditional GPUS, the company said. Tokens refer to data units processed by an AI model.
To remedy this, Nvidia will integrate several stages of the long processing sequence, such as video decoding, coding and inference – when AI models produce a result – in their new chip.
Investing US $ 100 million in these new systems could help generate $ 5 billion in symbolic revenue, the company said, as Wall Street is increasingly focused on the return on hundreds of billions of dollars in AI hardware.
Continues after advertising
The race to develop the most sophisticated AI systems made Nvidia the most valuable company in the world, running a dominant portion of the AI chips market with its expensive and first -rate processors.
(Reporterem in Arshees Bajwa Bajwa Bajwa)
