Broadcom and Qualcomm likely are producing more silicon.The only people who have any use for the foundries are Sam Altman and Elon Musk. If they have custom chip designs, they can make the chips without massive margins going to TSMC.
Broadcom and Qualcomm likely are producing more silicon.
AMD/Qualcomm/Nvidia are companies I wouldn't want to buy Intel. Any of these would lead to killing off a lot of consumer products.
I'd rather see Broadcom. The complaints against them seem to be that they'd increase prices. But that won't work with any of Intel's consumer products, they all have competition.
I would much rather see Nvidia merge with Intel. They actually give a damn about the technology, and won't destroy the company in a ketamine fugue
Ugh no, say goodbye to any open source software Intel has produced over the years. Plus that's when NV hardware starts to break on anything that isn't NV/Intel.I would much rather see Nvidia merge with Intel. They actually give a damn about the technology, and won't destroy the company in a ketamine fugue
Judging from how Twitter went, Elon will fire the teams supporting those open source projects and try to charge for access to them. So we're screwed either wayUgh no, say goodbye to any open source software Intel has produced over the years. Plus that's when NV hardware starts to break on anything that isn't NV/Intel.
He might. It's hard to say based on how he's handled SpaceX and Tesla. Tesla definitely isn't customer-friendly though (wrt software upgrades being non-transferable when selling used cars). Though he's probably only interested in the fabs.Judging from how Twitter went, Elon will fire the teams supporting those open source projects and try to charge for access to them. So we're screwed either way
Amazing
He also needs to sell Mars-hardened CPUs to the Mars colony once that gets going.Though he's probably only interested in the fabs.
He also needs to sell Mars-hardened CPUs to the Mars colony once that gets going.
/s
Is this the reason to buy Intel? Not Falcon Shores but a custom MI500 cluster on 14A. Would AMD say no if offered to build this many GPUs?Morgan: What’s the biggest AI training cluster that somebody is serious about – you don’t have to name names. Has somebody come to you and said with MI500, I need 1.2 million GPUs or whatever.
Forrest Norrod: It’s in that range? Yes.
Morgan: You can’t just say “it’s in that range.” What’s the biggest actual number?
Forrest Norrod: I am dead serious, it is in that range.
Morgan: For one machine.
Forrest Norrod: Yes, I’m talking about one machine.
Morgan: It boggles the mind a little bit, you know?
1.2 million GPUs is an absurd number (mind-boggling, as Forest quips later in the interview). AI-training clusters are often built with a few thousand GPUs connected via a high-speed interconnect across several server racks or less. By contrast, creating an AI cluster with 1.2 million GPUs seems virtually impossible.
I wonder which "podcasting bro" approached them about 1.2 million GPUs. 🤔AMD talks 1.2 million GPU AI supercomputer to compete with Nvidia — 30X more GPUs than world's fastest supercomputer
The best supercomputers in the world have less than 50,000 GPUs, how in the world is someone going to make an AI cluster with 1.2 million GPUs?www.tomshardware.com
Is this the reason to buy Intel? Not Falcon Shores but a custom MI500 cluster on 14A. Would AMD say no if offered to build this many GPUs?
I'm more curious which dumbass will give them enough funding for this. The podcast bros are just spending other people's money, who is going to pay the multiple billions to build and run a computer of that size? It's just utter stupidity.I wonder which "podcasting bro" approached them about 1.2 million GPUs. 🤔
Podcasting bro. So it would be Microsoft by proxy.I'm more curious which dumbass will give them enough funding for this. The podcast bros are just spending other people's money, who is going to pay the multiple billions to build and run a computer of that size? It's just utter stupidity.
So much energy and money being wasted on this nonsense.
Podcasting bro. So it would be Microsoft by proxy.
Musk doesn't seem like the AMD type based on Jensen's comments.He could afford it even after buying the foundry.Elon Musk set up 100,000 Nvidia H200 GPUs in 19 days - Jensen says process normally takes 4 years
The GPUs were all part of an xAI super computer.www.tomshardware.com
Microsoft should spend that money fixing their busted operating systemPodcasting bro. So it would be Microsoft by proxy.
More money for the Windows group? They're spending their budget deliberately ruining it instead. Why give them more?Microsoft should spend that money fixing their busted operating system
Is Intel mentioned in there ?500 billion to AI DatacentersTrump announces up to $500 billion in private sector AI infrastructure investment
President Trump announced billions in private sector investment by OpenAI, Softbank and Oracle to build AI infrastructure in the U.S.www.cbsnews.com