Nvidia has put the spotlight on its next generation GPU architecture, which is said to lay the foundations for the world’s first computational graphics processing units.
Known as Fermi, the new graphics architecture is built by an entirely new ground-up design. The tech holds three billion transistors as well as 512 shader processing cores, and is targeted to facilitate high-performance computing.
Nvidia says the GPU will deliver “breakthroughs” in both graphics and GPU computing.
“I believe history will record Fermi as a significant milestone,” said Dave Patterson, director of Berkeley university’s Parallel Computing Research Laboratory.
“Nvidia and the Fermi team have taken a giant step towards making GPUs attractive for a broader class of programs,” he added, offering his view in an official company statement.
The company’s CEO, Jen-Hsun Huang, said it was "completely clear that GPUs are not just graphics chips anymore.”
Key details, from speeds to costs, remain unknown for now.