Locator: 46139NVDA.
Nvidia NVDA ’s earnings were impressive but still not enough to push the stock higher. [Comment: incorrect analysis.]
However, the company did enough to show that rivals such as Advanced Micro Devices and Intel have a long way to go to catch up with regard to artificial-intelligence chips.
Nvidia executives didn’t quite dismiss the competition in the earnings call following its October-quarter results, but they stressed their position as the largest player for AI chips.
Their key to maintaining that looks to be a faster pace of technology development, with new releases now set to come every year instead of every two years.
“There is a fundamental reason why we accelerate our execution…because it fundamentally drives down cost,” said Chief Executive Jensen Huang on the call.
More from the link:That pace could be hard for AMD and Intel to match. Nvidia is set to maintain a market share of more than 85% in generative AI accelerator chips next year, according to Raymond James analyst Srini
Pajjuri noted that some customers might pause their spending with Nvidia to wait for the new B100 chips, set for mass production in the second half of next year.
However, that also means customers might hold off on buying AMD’s MI300 AI chip in the meantime, or Intel’s Gaudi 3, which is due next year.
Nvidia’s data center revenue is now projected to hit $65 billion in the fiscal year ending January 2025, more than double the projected combined data-center sales of Intel and AMD for about the same period.
“Despite a growing list of competitors, Nvidia remains the place to train one’s generative AI algorithms largely as a result of the CUDA [software] platform that Nvidia has spent almost 20 years building,” wrote independent analyst Richard Windsor, who publishes Radio Free Mobile.
My hunch: a year from now, some of us will have gone from the “magnificent seven” to the “fantastic four” : AAPL, TSM. NVDA, and ASML.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.