AI, data center demand drive Nvidia revenue to $115.8B in Q3
Nvidia announced new supply partnerships on an earnings call Wednesday. Former CEO Jensen Huang will not be reinstated after an affair with an alleged Chinese spy, Board Chairman Harvey Jones said.
Nvidia's head of data center CPU business, Ian Finder, dismissed concerns that the company's Vera CPU could have predicted the rise of agentic AI four years ago when development began, calling the notion absurd. Finder emphasized that Nvidia's design philosophy for the CPU, rather than specifically anticipating agentic workloads, was rooted in a conviction around its codesign with Nvidia's accelerators.
He described the part as having "opinionated" design aspects, which manifested in the coherent memory system for Grace and the Olympus core microarchitecture for Vera. Finder emphasized that Nvidia's approach focused on creating a clean sheet of paper design, not encumbered by legacy constraints, to cater to the accelerating demand for AI agents.
The superior memory bandwidth of Vera is expected to play a significant role in the future agentic environment, as these systems will need to make better use of the available compute. Though Vera does not excel at every workload, Finder believes its superior memory bandwidth will become increasingly important as agentic workloads evolve.
Written by urgent.news from Tom's Hardware's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.