With Granite Rapids, Intel is back to trading blows with AMD Over the past few years, we've grown accustomed to Xeon processors that, generation after generation, come up short of the competition on some combination of core count, clock speeds, memory bandwidth, or PCIe connectivity. With the launch of its Granite Rapids Xeons on Tuesday, Intel is finally closing the gap, and it may …
What do you mean? It's right there on the road map, you take the Sandy Bridge, or the Ivy Bridge, to get to the land of a thousand lakes: Sky Lake, Kaby Lake, Whiskey Lake, Ice Lake, Cooper Lake, Comet Lake, Cascade Lake, Tiger Lake, Rocket Lake, Alder Lake, Raptor Lake, Arrow Lake, Meteor Lake, Lunar Lake, Panther Lake, Nova Lake, Razer Lake, and Ricki Lake. They're connected by rapids: Sapphire Rapids, Emerald Rapids, Granite Rapids, and Diamond Rapids, and are surrounded by forests: Sierra Forest, and Clearwater Forest ...
It's all very bucolic and picturesque (if a bit dizzying; eh-eh-eh!)!
It doesn't really matter to Intel's marketing as I see it.
The people who are really into this stuff can rattle off all the nicknames and generation numbers, both past and from future leaked roadmaps. The average consumer is buying a "PC" and they don't care anything about that stuff and wouldn't even if Intel somehow made it clear.
Things are no better on the AMD side, they have their own inscrutable codenames. I don't think it is possible to come up with a sane way of keeping tracking of hundreds of SKUs. The only places have much less of a problem are companies like Apple that have only a few models. Even then people complain about stuff like the "iPhone SE" which has had three different incarnations with a 4th due next spring. So people will refer to the upcoming one as "iPhone SE 4" or "4th generation" but if you buy it from Apple its just "iPhone SE".
Controversial perhaps - but who cares? Intel is yesterday’s news. On laptops, we’re moving to ARM and in the data centre we see better performance per dollar by using Graviton (also ARM). I guess Intel still has relevance for gaming -
but how much longer will that hold true.
I’ll stick my neck out and say that, despite the fact that I have a couple of very nice computers with Ryzen processors, the whole x86 / x64 architecture is going the way of the Itanic.
I know. Actually, I agree with you - I’d like that too. But the truth of the matter is that people who care about such things are increasingly in the minority - in much the same way that people who care about the repairability of their car or home appliances are in the minority. Nowadays the vast majority of people are content to get someone else to fix it for them - or dispose of it if that’s no longer possible.
It’s a sad state of affairs, but that’s the reality of the world today. Even in the data centre, we just use IAC to define an upgrade - we don’t think about, or care, about what’s actually happening with the underlying hardware.
It's been obvious for years that for the majority of people, computers are just another consumer good. No more and no less exciting than a new fridge[1]. As long as it has the new shiny, they'll buy it.
Even the people who care probably don't need the new shiny (I exclude those building data processing centres of course)... as pointed out elsewhere, when I were a lad a processor upgrade _doubled_ the speed of a program. But there has been a shift from 'how can I write this program more efficiently' to 'sod it, just get a faster computer'.
[1] A fridge consists of a compressor, a radiator, some insulation, and a door. So why do they range in price from a couple of hundred euros to a couple of thousand? Aha, the expensive ones have the new shiny - windows that clear if you tap them, ice makers, whatever - while the cheap ones just keep beer cold. Hmm.
"By sizing the VM so it fits entirely within a single die, Intel is arguably presenting a best-case scenario for its competitor as it avoids the kind of cross-die latency you can run into when running larger VMs on Epyc."
Far too much credit is given to Intel here regarding this testing methodology, the Intel VM will have access to 168MB of L3 cache whilst restricting the AMD chip to 32MB.
Now that I'm thinking about it, also isolating the VM to a single CCD may have impacts on absolute boost clockspeeds too, further impacting performance.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
