Search
NVIDIA's Forthcoming Feynman GPU: Leveraging 1.6nm Tech and Exceeding 1000W Power Consumption
8 hour ago /
Read about 0 minute
Author：小编
NVIDIA is set to hold its Graphics Technology Conference (GTC) from March 16 to 19, during which it is anticipated to introduce at least two significant innovations. One of these is an LPU chip tailored for AI inference, designed to overcome the cost and efficiency limitations of traditional GPUs in inference operations. The other is the next-generation GPU architecture, named Feynman, following NVIDIA's custom of honoring eminent scientists in its architecture naming. Although prior roadmaps have suggested its integration with next-generation HBM memory, concrete details are yet to be released.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
