Monday, February 21st 2022

Intel Advancing 13th Gen Core "Raptor Lake-S" Launch to Q3-2022?
Intel is allegedly advancing the launch of its 13th Gen Core "Raptor Lake-S" desktop processors to some time in Q3-2022, according to a report by Moore's Law is Dead. It was earlier believed to be a Q4 launch, much like "Alder Lake" was, in 2021. The report predicts the debut of "Raptor Lake" in the desktop segment in Q3-2022 (between July and September), with certain mobile SKUs expected toward the end of the year, in Q4. The Core "Raptor Lake-S" processor is built in the existing Socket LGA1700 package, and is being designed for compatibility with existing Intel 600-series chipset motherboards with a firmware update.
The "Raptor Lake-S" silicon is built on the existing Intel 7 (10 nm Enhanced SuperFin) node, and physically features eight "Raptor Cove" P-cores, along with sixteen "Gracemont" E-cores that are spread across four clusters. The chip has additional cache memory, too. Moore's Law is Dead predicts that the "Raptor Cove" P-core could introduce an IPC uplift in the region of 8 to 15 percent over the "Golden Cove" core, while the chip's overall multi-threaded performance could be anywhere between 30 to 40 percent over "Alder Lake-S," on account of not just increased IPC of the P-cores, but also eight additional E-cores.
Sources:
Moore's Law is Dead (YouTube), VideoCardz
The "Raptor Lake-S" silicon is built on the existing Intel 7 (10 nm Enhanced SuperFin) node, and physically features eight "Raptor Cove" P-cores, along with sixteen "Gracemont" E-cores that are spread across four clusters. The chip has additional cache memory, too. Moore's Law is Dead predicts that the "Raptor Cove" P-core could introduce an IPC uplift in the region of 8 to 15 percent over the "Golden Cove" core, while the chip's overall multi-threaded performance could be anywhere between 30 to 40 percent over "Alder Lake-S," on account of not just increased IPC of the P-cores, but also eight additional E-cores.
52 Comments on Intel Advancing 13th Gen Core "Raptor Lake-S" Launch to Q3-2022?
Go ahead and cram 16 P-cores on the same monolithic die. Good luck fitting that let alone cooling that. Intel goes MCM? Cue the "glue" comments that definitely aren't overused. People still can't stop complaining and whining. My original point, alive as it always has been.
The E-cores managed to get the 12900K above and beyond the 5950X, so what is the problem exactly? They function really well, manage background tasks perfectly, are efficient, and also strong when paired together in a multithreaded task, so what is the problem exactly?
I wonder why AMD is having some cores imitate E-core behavior by being lower clocked on Zen 4 if E-cores are so bad. Hmm.
This is my final response on the matter, read my signature if you care to know why.
Currently im on 3800X. I skipped regular 5000 series and will jump to 5800X3D. Then wait out both Raptor Lake and and Zen 4 launches. Maybe when Meteor Lake and Zen 5 will arrive in 2023 i might consider upgrading. By then we should hopefully have unquestionably faster DDR5 at normal prices and AM5 teething issues would have been solved.
Yeah I'm sure it's light years faster than that one :laugh:
Seriously, each generation seem to last less than 6 months, with few upgrades, far for AMD (except for ST).
They should take a moment, and leap in with a fierce new upgrade, a solid one.
Not an asthmatic new Gen Core every 6 month that nobody will buy.
Yeah guessing Intel hasn't noticed a mineral/ parts shortage as everyone else has ?
In the future this kind of architecture can be great - like imagine having the main game rendering on P cores and the static UI elements using E cores - but that's not today. Today it's only the way Intel found to be able to match the core count AMD was offering since MCM and EMIB aren't ready for prime time yet.
We can mock EMIB glue just as we mock Ryzen infinity latency until the tech proves it self, just like E cores. When things stop being just marketing ploys there will be all the reasons to be happy, until then E cores are a marketing trick, they aren't necessarily useless or crap but they aren't any marvel of engineering either.
People don't want this though, they want 32 cores on their desktop so they can have 20 idle doing nothing, the same as people having 32 or 64gb of ram on a gaming rig, so they can have 20 or 50gb doing nothing. Except it sure looks good having tons of boxes when you show your mates your task manager CPU display.
I don't personally think Intel did it to match AMD. Why Amd made a desktop chip with 16/32ht cores is a mystery, as they certainly are not needed for a normal user. See how Intels 8P cores matched the 16/32 of AMD, can't count the E core can we as they are useless just to make up the numbers. Shows how good them 16 cores of AMD must be though. Imagine 16 Intel P cores, Power might be high but it would be hard for AMD to match it with any current desktop CPU.
I am happy with the 8P cores, i don't feel my PC needs any more, others seem to crave them though. R23 scores re not the be all.
Anyway... I'm not crying for more P-cores. 8 are perfectly fine for me. All I'm saying is that I can't see the point of 16 E-cores (or 16 any cores) on a mainstream desktop platform. All that AMD's 3950X and 5950X did imo, is make their HEDT platform pointless for a lot of people. If the counter-argument is that Intel's HEDT segment hasn't seen an update lately and Core i9 is the new HEDT in a mainstream socket, just like the 5950X with AMD, I can accept that.
I'm not crying for more P cores either the 8 P is as good as AMD's 12 core 5900x. I just accept the fact my 12700k has E cores, they do something. It just bugs me people saying they are slow, error cores, useless etc as they certainly are not. They are also not atom cores by any means.
Chip is not nearly as big as hedt though guess it's early though they'll be as big as 2066 chips soon enough but oops another new freaking socket :banghead:
Markting would have us believe "E cores are great for low priority things", well they aren't bad, but aren't great either, just a new marketing bulletpoint. People want the best, regardless if they need it or not. If I can have 16 or 12 full fat P cores why would I choose to have 8 P + 4 E or 8 P + 8 E cores? Sometimes I might, but the simple 16 or 12 P cores sound better (I know I know, the 8P + 8E can outperform the 12P). Well then the markting worked. Why would they bother with E cores after years of the "you don't need more cores" argument!? Lots of normal users need 16 cores, just depends on what you consider normal. If that wasn't the case, Intel wouldn't also be increasing the number of cores. R23 is meant to represent 1 type of workload, that happens to be regarded as a good-ish way to measure saturated performance. I'm a software developer, R23 results don't really translate but I look at a high score as my code will compile faster. They're not everything, but better is better.
The choice is not number of cores but performance surely. If 8 out perform 12 (or whatever) i would take the 8 every day.
Intel is increasing the number of cores, just not P ones, which seems to be why some people are pissed at Intel, they want P cores as more is better surely. If people need more cores does it really matter if less out perform more? surely not.
As for r23, even stock, my CPU easily beat a 5900x in ST and MT so if you need to compile stuff then the higher scores are better, even with less cores. Sometime more is not always better.