this post was submitted on 12 Dec 2024
249 points (93.1% liked)
Not The Onion
12551 readers
1642 users here now
Welcome
We're not The Onion! Not affiliated with them in any way! Not operated by them in any way! All the news here is real!
The Rules
Posts must be:
- Links to news stories from...
- ...credible sources, with...
- ...their original headlines, that...
- ...would make people who see the headline think, “That has got to be a story from The Onion, America’s Finest News Source.”
Comments must abide by the server rules for Lemmy.world and generally abstain from trollish, bigoted, or otherwise disruptive behavior that makes this community less fun for everyone.
And that’s basically it!
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
This is assuming there isn't some gold in the pipeline. Timeline on a new CPU design is about 8 years from first drawings to actual silicon hitting media test benches, meaning whatever was started in 2019 and 2020 could be absolutely killer and just cooking to perfection in the R&D oven...assuming R&D was kept sufficiently funded and the engineering talent retained to see such a process through
Their CPUs are mostly fine now. The “small” cores are very competitive in servers because they are so small for their perf, and TBH that desktop drama was marketing clocking the CPUs way too high to squeak out 4% more benchmark performance.
It’s… everything else that’s the problem, as work is increasingly shifting away from CPUs. They are totally screwed if they cut funding for Arc, in particular, or if they don’t secure any real fab customers.
Oh yeah real talk they've got some killer datacenter chips, be that networking, CPU or GPU. They're continuing to work on bleeding edge technologies for hyperscalers, and they've got no shortage of insane potential. But when they release two generations of desktop processors with hardware bugs it really puts a heck of a stain on such a stellar portfolio and makes it a lot easier for enterprises to look at AMD for their datacenter and client processors (especially when they're absolutely killing it like they have been in both segments!)
Eh Intel’s data center GPUs suck. Gaudi was okay, but gained no critical mass and is being phased out, Xe-HPC is being phased out and was hardly used anywhere, even Falcon Shores keeps getting delayed and looks to be in trouble going by statements about focusing on “consumer inference.” They seem dead in the water here, which is very worrying.
The MI300X is actually good, but AMD totally blew it by ignoring a few glaring software issues and not seeding development with consumer GPUs, hence it’s not gaining much traction. The MI300A (the big APU) basically isn’t available or cost effective in any cloud instances.