US chip designer and software program specialist Nvidia on Tuesday mentioned that it needs to fulfill the quickly rising demand for computing energy for synthetic intelligence (AI) with a brand new era of its chips.
Chief govt officer Jensen Huang mentioned at Nvidia’s developer convention GTC in San Jose, California that the corporate’s next-generation Vera Rubin chip was scheduled to be launched in autumn 2026.
Vera Rubin and the additional improvement of the Blackwell system, introduced for this yr, are anticipated to drastically cut back the prices of working AI software program in comparison with earlier know-how.
Nvidia chips have grow to be a key know-how for synthetic intelligence.
Industry giants akin to Google and Meta, together with AI start-ups like ChatGPT creator OpenAI, depend on Nvidia’s chips to coach synthetic intelligence techniques.
This place allowed Nvidia’s business to develop explosively in recent times.
Huang additionally tried to allay buyers’ considerations that the world may get by with much less AI computing energy in future – and that expectations of Nvidia’s future business may subsequently be too excessive.
He mentioned that the world as a complete is transferring in direction of producing solutions from scratch with the assistance of synthetic intelligence, as a substitute of retrieving saved solutions. In explicit, the brand new AI fashions that may construct a sequence of reasoning step-by-step to unravel an issue are power-hungry.
Overall, you want 100 occasions extra computing capability than was assumed only a yr in the past, mentioned Huang.
Source: www.anews.com.tr