Given* how they're upping the clocks across the board, I would dare saying that cache difference is not going to matter a whole lot for those lower end parts. Well, as I said, performance is what will matter, so we'll see.
We will have to see when they release but it doesn't make any sense to me.
Why would intel rebrand old dies if they can just launch only the K line-up for starters and then wait for the AL cpus to sell out (enough) to release the non-k?
Also what? The more expensive AL K models are the ones that sold so much better then the cheaper non-k ones? How does that make sense?
Or do they think that the RL K will sell so much better even though the AL K didn't?
Given* how they're upping the clocks across the board, I would dare saying that cache difference is not going to matter a whole lot for those lower end parts. Well, as I said, performance is what will matter, so we'll see.
Higher clocks do you little good without the cache hit rates to keep cores fed. Most of Alder Lake's IPC gains were mainly from increasing the L3$ size by 50+% across the board and most of Raptor Lake's IPC gains will come from doubling L2$. Clock bumps are extra performance on top of that, not a substitute.
I suspect Alder Lake mainly failed because almost everyone who needed a new PC during COVID got 10th-11th gen and won't be upgrading again any time soon. Corporate office sales are on a 3-4 years cycle and may take ~10 years to even back out while consumer sales may take 7+ years to do the same.
Meteor Lake is when the market may be on its way to return to relative normalcy with the first post-covid corporate and institutional computer fleet refresh.
But they could never get the production cost down enough to be able to sell it at prices that gave them profits and made it attractive to customers. Even at the ridiculously high prices they were selling it at they lost billions.
Micron, someone that knows a helluvalot about memory production saw the writing on the wall and dumped it two years before Intel finally admitted it was a dead end.
Dumping Optane will, unfortunately, be Pat G's sole positive contribution to Intel before he is dismissed.
I suspect Alder Lake mainly failed because almost everyone who needed a new PC during COVID got 10th-11th gen and won't be upgrading again any time soon. Corporate office sales are on a 3-4 years cycle and may take ~10 years to even back out while consumer sales may take 7+ years to do the same.
That would make sense if the covid sales boom only lasted one year, but it has been pretty much 3 years of higher sales, and for intel it has been another two years of the same high sales before covid,
Intel has had double the net income ever since 2017. two whole years before covid ever started.
And also we have zero real numbers on alder lake sales, the sales might have been terrible or great or normal.
We will have to see when they release but it doesn't make any sense to me.
Why would intel rebrand old dies if they can just launch only the K line-up for starters and then wait for the AL cpus to sell out (enough) to release the non-k?
Also what? The more expensive AL K models are the ones that sold so much better then the cheaper non-k ones? How does that make sense?
Or do they think that the RL K will sell so much better even though the AL K didn't?
Those are good questions and I share them, since I also find it curious.
Other than simplifying the argument by saying "AMD just sold a lot more", I think it's not the only reason if the rumored information is true. Could it be they expected the trend to still be upwards and were hit with the slowing process way before they expected it? Maybe higher end parts sold way less than what they originally estimated and have a surplus of parts? OEM got ahead of the slowdown and just halted purchases/orders way in advance?
This definitely is linked to the slowdown of the market, somehow. So maybe that's just it?
It would've been an entirely different picture if Intel made Optane attractive from a price standpoint. Intel's greed sent Optane to its grave, unfortunately.
I'm really not, and I think the majority of still sane and sober people ain't either, but cheer that Intel finally decided to kill Optane – At last a mature decision.
No, it really wasn't, like at all. It was a miserable excuse of a product and for sure never should've entered actual product-stage.
Yet, the constant cheering over it of a gazillion of Intel's own f@nb0ys, prevented killing it out of reasoning and sanity.
No offense here personally, but there it is again. The feelings™ .. You're »sorry« that it died. Thus, feelings are involved again. Though, feelings are the single-worst advisors for anything. Just like the saying goes, »Angst is a bad advisor.«, are feelings in general the worst advisors when rational decision have to be made. Decision over hard cash and survivability, especially if such decisions involve a business with +100K employees like Intel is. That's how big companies are primed for a sudden downfall or slow death.
Intel is exactly that, and the Optane-endeavor (among others) showed exactly that again: Intel is wasting billions over feelings!
Optane never should've left the drawing board, since it was a technology which was never economically viable to manufacture, as the actual price-tag (with forward-charged added profits) would have been so damn sky-high, already out-weighting its cost-benefit-ratio by a mile, that it was basically plain unmarketable. Well, apart from the fact, that its very use-cases were nigh existent to purely academic to begin with and its actual procurement hard to justify at each and every price-tag.
It was a fancy idea, to philosophise and fantasise about for a minute or two on a nice coffee-break, but that's about it.
It would've been an entirely different picture if Intel made Optane attractive from a price standpoint. Intel's greed sent Optane to its grave, unfortunately.
It wasn't profitable on its own anyway (Optane never netted a single dime of profit for anyone involved but instead huge losses of billions ever since) and was only tenable (and especially affordable) as a introduction-device and vendor lock-in into the pricy Intel-world, by working as the prominent Xeon-kicker as a unique selling proposition when bundled with their Xeons – The outrageous Xeon-profits would have had made it affordable.
Edith wants to note here: The moment Micron got it altogether and finally took a sharp look on the balance-sheet, Micron knew for sure (whilst they likely always assumed it since years), that it wasn't any economically viable to manufacture and outright impossible to afford as a stand-alone-product. Micron must have known, they manufactured Optane for Intel.
And you can bet your sit-upon on the fact, that Intel sold their stake of the joint-venture IMFT (Intel-Micron Flash Technology) to Micron with largely cooked books and hid billions of losses by shifting losses back into their own balance-sheets (which were largely profitable by then back then)!
Remember, the stake was sold to Micron with already almost a billion of debts attached to it which Micron had to pay. The $1.4B-takeover of the Intel-stake was actually worth only about $400M (for the plant in Lehi, Utah and the respective tooling), the rest were debts Micron had to pay.
The moment Micron had insight into the financial side of thinks, thus the actual books (which Intel always managed ever since; Micron exclusively managed the manufacturing side of things ever side, that was the actual division of tasks), they instantly knew, Intel was pretending Optane's profitability all along since 2006! That's why Micron immediately pulled the plug out of 3D XPoint aka Optane the moment they saw the actual deep-red financial numbers behind it in bold, capital letters at the bottom of it.
Also keep in mind, that even from 2019 onwards, where Micron only manufactured 3D XPoint for Intel and had Intel as the one and only customer, Micron already made almost half a billion of losses each running year (~ $400M/yr) due to massive underutilisation. Since surprise, suprise, Optane wasn't selling at all, no matter how often Intel claimed it would be profitable and selling like hot cakes. Intel lied, what a suprise.
Remember: After all, Intel is famous for their everlasting financial engineering and their infamous cross-subsidisations to hide huge losses by shifting around lossy departments under highly profitable divisions.
The problem just is, that Intel was so slow to bring it to market and hit any whatsoever time-line (as always), that it reached the market by the time, AMD finally had their Ryzen again. Tough luck selling some outrageous overtly expensive USP-device as the Xeon-kicker, when the competition is starting to undercut you by a mile with less expensive yet more performant options a buyer can chose from.
It never should've left the drawing board, nevermind trying to create a product out of it for aforementioned reasons. Especially trying for literally years to forge a product over a fancy theory and moot use-case while mindlessly pouring billions into it over hurt feelings of false pride.
Yet Intel always tried to create use-cases where none were existing (to justify its unjustified existence beforehand) and poured literally billions into Optane, to maintain it into life (by selling it way below manufacturing-costs), when it never should've lived as a independent product anyway in the first place and was even highly questionable as a USP-vehicle.
Though, it's coming from Intel after all. That one company, where the divisions and departments are somehow allowed to bring to market a product literally no-one asked for, has basically none whatsoever greater use-case and for sure no market to be sold to independently and isn't affordable as a independent product on its own. Yet it gets pushed through mindlessly due to big egos and wounded pride. Same story happened to Larrabee, Xeon Phi, Itanium or other failed Intel-projects before. Billions for naught.
Yes, please do because your whole rant of this post was only your feelings on this matter with not a single piece of actual data backing anything up.
Were are the sales numbers that show how bad it really was?
Were are the evaluations intel did before they made the product to show how much if any market there was?
Were are the actual business plans?
See without any of that all you are doing is expressing your feelings.
If optane ram modules would have replaced a big portion of storage it would have benefitted everybody from people with mobile devices over simple home users up to huge server customers.
to bring to market a product literally no-one asked for, has basically none whatsoever greater use-case and for sure no market to be sold to independently and isn't affordable as a independent product on its own.
Do you know how the quote "Nobody will ever need more than 640k of RAM " came to be??
Because anything above that was freakishly expensive at the time and nobody thought that anybody would ever spend so much money on getting more than 640k.
All of the things you say here are 100% applicable to above 640k back in those days, and here we are now with 16Gb not being enough for many people.
In other words:“Please let me at least look and feel superior by insulting you personally with hypocritical ad hominem-attacks,
when I'm aware that I was outargued to make your post look less irrelevant in labelling it as a rant …“
If anyone needs actual data to back this up (thus data which hasn't been already in the news for weeks), months or even years, I don't know what to tell you here …
Optane never made a dime of profit for everyone involved and only netted a loss for everyone involved, it wasn't selling like Intel claimed it to be and that's exactly why it was knifed.
That are not my feelings, like you put it, but actual facts being in the news even here on Tom'sHardware. Remember that Intel still sits on a 2-year stockpile of excess-inventory of it!
Plain and simple, Optane was a loss maker for Intel from start to finish and even Micron lost hundreds of millions if not already billions on it due to massive underutilisation (since it wasn't selling, like Intel claimed it would) and by paying for Intel's losses when they overtook the stake in the joint-venture's share from Intel. It was a $1B-loss directly attached to it by that transaction alone.
Were are the sales numbers that show how bad it really was?
Were are the evaluations intel did before they made the product to show how much if any market there was?
Were are the actual business plans?
See without any of that all you are doing is expressing your feelings.
That has nothing to do with feelings you fool!
Intel was losing money on it from day one and just covered up the losses for years with cross-subsidisation via sneaky financial enggineering.
If optane ram modules would have replaced a big portion of storage it would have benefitted everybody from people with mobile devices over simple home users up to huge server customers.
It couldn't have any greater market-availability since the surplus-charge was so absurdly high and excessive over anything else and didn't even was largely better in the majority of use-cases. If you think Optane would've been able to become mainstream, that POV reeks of pure delusion.
Facebook
Twitter
Instagram