Intel to Go Ahead with "Meteor Lake" 6P+16E Processor on the Desktop Platform?

[A&P211]

Intel really seems to love these stupid and useless cinebench accelerators...
i don't want ANY E-Cores in my DESKTOP CPU. they make sense for devices running off a battery but they are completely useless in a normal PC.

give me 16 P Cores, downclock them to keep heat and power consumption tame enough for the average user and let me overclock it back to it's limit with a high end cooler.

The e-cores seem to be mostly useless on laptops.

 

[evernessince]

If you're going to make a comparison about E cores using a chart with the 1st generation Alder Lake E cores rather than Raptor Lake E cores really isn't the best place to start. That's even more true when talking about 3rd generation E cores making even further refinements to E cores designing. It's a bad comparison choosing to look at the older design especially when there were big changes to the design with Raptor Lake for the better like individual E core multipliers and voltage controls instead of clusters of four. Effectively it's higher resolution control over the multiplier setting and voltages which leads to better sloping of parameter settings for efficiency and performance.

Intel made the right changes to E cores with Raptor Lake as a whole it was good step forward. If anything they just pushed P core frequency too damn far at peak. It's been pointed out that it's fairly easy to correct most of the efficiency problem w/o sacrificing overly heavily. To be fair the AM5 7800 vs AM5 7800X is quite a difference as well and shows AMD pushed the 7800X a lot further and could've had a lot better efficiency as well from the get go. Neither company were perfect on that end.

Raptor lake E-Cores made a good step forward? Power consumption is worse than the 12900K and so is efficiency for the performance provided:

And no, it's not possible to just tune down the performance of 13th gen without loosing significant performance:

 

[Chrispy_]

Either these Crestmont E-cores are half the size of previous gen, or those are images of 6P+8E chips.

 

[Vario]

Yeah I misspoke, meant #cores in the case of AMD.
Still though - 12T vs 12T; so whichever way it goes, on a straight bench that can use every core thrown at it, E cores don't extract a real advantage. 'Tuned to meet' - rather, I'd say, E cores are tuned to meet a TDP target to ensure the chips don't straight up burn in hell. Of course, this happens on the AMD side too, except they're a lot smarter about it now.

But the fact still remains, that on an AMD CPU, you can use every core for every task without scheduling or other shenanigans. The fact is also that in any full-blown, unlimited load the Intels go straight into crazy land wrt power usage, while AMD's recent 7950X3D peaks at half TDP of a top- and even subtop- Intel part.

So that's where we see the real thing. In any limited scenario, Intel can keep up. Remove limitations and the AMD parts deliver peak performance at fantastic efficiency, and the Intel parts start showing their true, excessive TDPs. And the difference it so seems, is mind blowing.

So the matchup is pretty much equal die space for similar performance, but twice the power usage at peak due to 'Efficient Cores'. Well played, Intel, well played indeed, gullible consumers buying the marketing. And why? So Intel can 'keep up'. Yeah, that Big little sure is a winner on the eternally rehashed Core CPUs, go go.

View attachment 285704View attachment 285705

Efficiency chart is hilarious, even. Tell us again Intel didn't get stuck on quad, maybe hexacore since forever; the only parts that have any semblance of effiency in the current day are low core count parts. Apparently mix&match your old crap to make ends meet doesn't quite suffice against actual technological progress

View attachment 285706


You either have sufficient core count or you don't, its that simple, and it always has been. But then again, there's a lot of blundering going on @ Tweakers, be wary taking those reviews too seriously. They're Hardware.info level now - bottom barrel, up to and including straight up wrong results. I've had my share of experiences. Even prior to HWInfo invading to take over the abysmal review quality, they 'oopsied' on for example The Witcher 3 testing with Hairworks on. Yes you read it right. It took some heavy complaining from this person to correct that nonsense. Reviewers are liable to speak for the very thing they spoke against less than a month ago, etc. Its a mess.

Also, interesting that you do full blown downloads in background while gaming, that'll be some enjoyable ping!

I can't apologize for the 12900KS's terrible power consumption out of the box, certainly they could have done better if they had a more competitive process (smaller node).

A lot of the issue with Intel is that Alder lake is clocked higher than it should be, simply for the sake of generating a lead or a tie on benchmarks. I found limiting my 12900KS to 5.0 and undervolting it dropped the wattage consumed by 100+ W depending on application, and the temperatures don't much exceed 60C in games. If I ran it out of the box, it would be 110C and 300 watts.

 

[THU31]

If the 14600K gets 6 P-cores then this could still be an attractive platform with an upgrade path to Arrow Lake.

The PCI-E connectivity looks very appealing.

Meteor Lake should be limited to i7 SKUs at the top. I don't believe they'd be dumb enough to release an i9 with 6 P-cores, even if the Cinebench score could somehow top the 13900K.

 

[lemonadesoda]

It means I could have many poorly threaded Adobe apps open at once rather than 1 or 2.

>> I could have many poorly threaded Adobe apps open at once LIKE
>> I would also need terabytes of RAM for many Adobe apps open at once NO LIKE

 

[MrMilli]

Yeah I misspoke, meant #cores in the case of AMD.
Still though - 12T vs 12T; so whichever way it goes, on a straight bench that can use every core thrown at it, E cores don't extract a real advantage. 'Tuned to meet' - rather, I'd say, E cores are tuned to meet a TDP target to ensure the chips don't straight up burn in hell. Of course, this happens on the AMD side too, except they're a lot smarter about it now.

But the fact still remains, that on an AMD CPU, you can use every core for every task without scheduling or other shenanigans. The fact is also that in any full-blown, unlimited load the Intels go straight into crazy land wrt power usage, while AMD's recent 7950X3D peaks at half TDP of a top- and even subtop- Intel part.

So that's where we see the real thing. In any limited scenario, Intel can keep up. Remove limitations and the AMD parts deliver peak performance at fantastic efficiency, and the Intel parts start showing their true, excessive TDPs. And the difference it so seems, is mind blowing.

So the matchup is pretty much equal die space for similar performance, but twice the power usage at peak due to 'Efficient Cores'. Well played, Intel, well played indeed, gullible consumers buying the marketing. And why? So Intel can 'keep up'. Yeah, that Big little sure is a winner on the eternally rehashed Core CPUs, go go.

View attachment 285704View attachment 285705

Efficiency chart is hilarious, even. Tell us again Intel didn't get stuck on quad, maybe hexacore since forever; the only parts that have any semblance of effiency in the current day are low core count parts. Apparently mix&match your old crap to make ends meet doesn't quite suffice against actual technological progress

View attachment 285706


You either have sufficient core count or you don't, its that simple, and it always has been. But then again, there's a lot of blundering going on @ Tweakers, be wary taking those reviews too seriously. They're Hardware.info level now - bottom barrel, up to and including straight up wrong results. I've had my share of experiences. Even prior to HWInfo invading to take over the abysmal review quality, they 'oopsied' on for example The Witcher 3 testing with Hairworks on. Yes you read it right. It took some heavy complaining from this person to correct that nonsense. Reviewers are liable to speak for the very thing they spoke against less than a month ago, etc. Its a mess.

Also, interesting that you do full blown downloads in background while gaming, that'll be some enjoyable ping!

I don't know if you're Dutch or not but Tweakers is not "bottom barrel". Please ... They're average and any website has the occasional mistake. That graph I posted is still valid.
It's not only about core count. The e-cores have their own cache pool which also helps with multitasking.

Newest review:

PS: ping is great but I don't have to worry with my 500Mbit internet.

 

[Minus Infinity]

I'm a little surprised there is so much push back about E cores sure if intel released a 10/12 Pcore cpu I'd be on Alderlake/raptorlake but the majority of people I personally know that have 12/13th gen intel are very happy with their systems same with people I've personally done AM5 systems for... Don't get me wrong I'm not thrilled about meteor lake being 6P cores but I'd rather see benchmarks before deciding if it's not for me. My biggest worry if it does suck we will just get amd pricing 6 cores at 300+ and 8 cores at 400+ again which is worse than intel using E cores imo.

I'm curious why you wouldn't wait fro Arrow Lake anyway? So much new stuff Intel is bringing to the table: new node, new chiplet design and FPGA. I'm guessing this is why the deskptop is posing problems and we aren't seeing 8P at all even for mobile. Arrow Lake's refined node is a much bigger improvement and we get 8P cores showing another large IPC uplift over Meteor Lake. I realise we have to trust Inte's roadmap timeline, which is a stretch, but I would not bother updating from Raptor Lake or Zen 4 to Meteor Lake for desktop assuming it even appears. Arrow lake is the first exciting thing I seen from Intel in a long time.

The other fly in the ointment is whwther the much rumoured Zen 5 hybrid architecture is real with 5c cores making an appearance. I would think AMD would want to make higher core count models while keeping power in check. The 4c Bergamo cores would be much stronger than current e-cores and I expect 5c cores to still be much more powerful than Crestmont e-cores in Meteor Lake.

Now maybe Meteor Lake will be a surprise but I still consider it the test bed for Arrow Lake and would avoid no matter how good.

 

[Vayra86]

I don't know if you're Dutch or not but Tweakers is not "bottom barrel". Please ... They're average and any website has the occasional mistake. That graph I posted is still valid.
It's not only about core count. The e-cores have their own cache pool which also helps with multitasking.

Newest review:
View attachment 285746

PS: ping is great but I don't have to worry with my 500Mbit internet.

We'll agree to disagree about Tweakers, I've seen far too many mistakes and straight up marketing parroting over the years. They are definitely bottom barrel, and heavily sponsored from time to time, too. Neutral... dream on.

 

[MrMilli]

We'll agree to disagree about Tweakers, I've seen far too many mistakes and straight up marketing parroting over the years. They are definitely bottom barrel, and heavily sponsored from time to time, too. Neutral... dream on.

I don't understand why you're changing the subject into discrediting Tweakers. That's very politician like. Don't like the fact, discredit the source. I don't want to go into your rabbit hole.
What would be better from you, is to present me with the opposite (the grown up way as it's called). Show me another review that does a multi-tasking gaming test (really not many out there) where AMD doesn't fall behind.

The fact is that those stupid e-cores bring along their own private cache which helps a lot when you use your computer as it's actually intended (ie. multi-tasking). This lowers cache thrashing of which AMD is clearly suffering in that test. The addition of X3D clearly shows that.
As you so eloquently put "You either have sufficient core count or you don't, its that simple, and it always has been" is a very primitive and brute force way of solving a more complex problem. Advanced architectures such the Fujitsu A64FX show how a well thought-out memory hierarchy overcomes the need of MOAR CORES.

 

[AnotherReader]

Advanced architectures such the Fujitsu A64FX show how a well thought-out memory hierarchy overcomes the need of MOAR CORES.

The A64FX doesn't show anything of the sort; it shows what a well designed vector implementation and ISA can do with a memory hierarchy capable of feeding the beast. Given its low clock speed, you wouldn't be singing hosannas about it if you were to use it as a desktop CPU. It also has a lot of cores for its time: 52 cores.

 

[MrMilli]

The A64FX doesn't show anything of the sort; it shows what a well designed vector implementation and ISA can do with a memory hierarchy capable of feeding the beast. Given its low clock speed, you wouldn't be singing hosannas about it if you were to use it as a desktop CPU. It also has a lot of cores for its time: 52 cores.

It was just an example obviously and well said, feeding the beast, which is basically what every CPU suffers from. Fujitsu alleviates this by its cache hierarchy, HBM2 RAM and Tofu interconnect. Fujitsu managed to keep those 48 cores fed, which other ARM server CPU's failed in.
I didn't mean that A64FX or SVE were any use for the desktop. I used it as an example that bruce-force is not always the answer.

 

[Vayra86]

I don't understand why you're changing the subject into discrediting Tweakers. That's very politician like. Don't like the fact, discredit the source. I don't want to go into your rabbit hole.
What would be better from you, is to present me with the opposite (the grown up way as it's called). Show me another review that does a multi-tasking gaming test (really not many out there) where AMD doesn't fall behind.

The fact is that those stupid e-cores bring along their own private cache which helps a lot when you use your computer as it's actually intended (ie. multi-tasking). This lowers cache thrashing of which AMD is clearly suffering in that test. The addition of X3D clearly shows that.
As you so eloquently put "You either have sufficient core count or you don't, its that simple, and it always has been" is a very primitive and brute force way of solving a more complex problem. Advanced architectures such the Fujitsu A64FX show how a well thought-out memory hierarchy overcomes the need of MOAR CORES.

I'm not changing the subject - neither am I saying that those specific results are 'wrong' or somehow twisted. What I am questioning, is 'what's the actual point' of the results you showed. As I pointed out earlier, the overwhelmingly vast majority of people don't download and game online at the same time. I've done so on a pretty good connection and its not pretty. Regardless of bandwidth, your ping/latency will suffer. Maybe life's different today, but still- why. #firstworldnonissues comes to mind.

Similarly; we're looking at 500 FPS while streaming and gaming on a CPU that uses twice as much power as its nearest competitor. So now you have a CPU burning through north of 200W continuously to display 500 FPS and stream a game... at 60 FPS... and heavily compressed. I'm sure there is a world where this is considered absolutely fantastic, but it isn't mine.

As far as your other points about E-cores, absolutely. I'm not going to deny there IS a niche for this architecture, I even acknowledge it, just not in the same way as you: Intel needs and uses these E-cores not to be better or faster, but to keep up by using and still refining what is by now ancient tech. What's happening now is that they're pushing higher E core count because even that wasn't enough, after multiple changes to how we should look at specs, turbo, etc etc ad infinitum, and with power usage soaring to unseen heights. The 7950X3D absolutely destroys everything these E cores produce, private cache or not - and that's just by adding cache to a chiplet CPU. No wild nonsense babble about 'efficient' cores, but just cores and more cache; and they produce the most stellar perf/w number we've seen historically, to date, apparently now not only in gaming but everywhere.

My main issue with E-cores is that they allow Intel to essentially stagnate yet again, and the only purpose they truly serve is marketing - core counts matter - while a competitor shows none of that BS is actually required or even offers a real advantage - Intel chips are just as big, more power hungry, and not faster.

It was just an example obviously and well said, feeding the beast, which is basically what every CPU suffers from.

This is part of my point above too.

 

[dyonoctis]

I'm not changing the subject - neither am I saying that those specific results are 'wrong' or somehow twisted. What I am questioning, is 'what's the actual point' of the results you showed. As I pointed out earlier, the overwhelmingly vast majority of people don't download and game online at the same time. I've done so on a pretty good connection and its not pretty. Regardless of bandwidth, your ping/latency will suffer. Maybe life's different today, but still- why. #firstworldnonissues comes to mind.

Similarly; we're looking at 500 FPS while streaming and gaming on a CPU that uses twice as much power as its nearest competitor. So now you have a CPU burning through north of 200W continuously to display 500 FPS and stream a game... at 60 FPS... and heavily compressed. I'm sure there is a world where this is considered absolutely fantastic, but it isn't mine.

As far as your other points about E-cores, absolutely. I'm not going to deny there IS a niche for this architecture, I even acknowledge it, just not in the same way as you: Intel needs and uses these E-cores not to be better or faster, but to keep up by using and still refining what is by now ancient tech. What's happening now is that they're pushing higher E core count because even that wasn't enough, after multiple changes to how we should look at specs, turbo, etc etc ad infinitum, and with power usage soaring to unseen heights. The 7950X3D absolutely destroys everything these E cores produce, private cache or not - and that's just by adding cache to a chiplet CPU. No wild nonsense babble about 'efficient' cores, but just cores and more cache; and they produce the most stellar perf/w number we've seen historically, to date, apparently now not only in gaming but everywhere.

My main issue with E-cores is that they allow Intel to essentially stagnate yet again, and the only purpose they truly serve is marketing - core counts matter - while a competitor shows none of that BS is actually required or even offers a real advantage - Intel chips are just as big, more power hungry, and not faster.

You say stagnate, I'll say "preventing a problem from becoming even bigger than it would if they listened to internet engineers". From puget tests, The latest XEON are really not that impressive even compared to zen 3 TR, they are losing in perf/watts even at a similar core count...against a 2 years old architecture. Even with their gigantic R&D, it seems that intel hasn't figured out how to make cores that are efficient and fast.

And while e-core reach their limits at the high-end, they are definetly a problem for AMD on the mid-range. The core i5 are way too good for what they cost compared to the competition. Unless they manage to pull a miracle, going back to a 6 core i5 will actually be a downgrade in many scenario compared to what they are now.

 

[Vayra86]

You say stagnate, I'll say "preventing a problem from becoming even bigger than it would if they listened to internet engineers". From puget tests, The latest XEON are really not that impressive even compared to zen 3 TR, they are losing in perf/watts even at a similar core count...against a 2 years old architecture. Even with their gigantic R&D, it seems that intel hasn't figured out how to make cores that are efficient and fast.

And while e-core reach their limits at the high-end, they are definetly a problem for AMD on the mid-range. The core i5 are way too good for what they cost compared to the competition. Unless they manage to pull a miracle, going back to a 6 core i5 will actually be a downgrade in many scenario compared to what they are now.

I'll happily wait and see what AMD does with the Big little idea. But maybe they're concluding they won't even need it.

Similarly... Intel on chiplet... bring it on.

 

[ThrashZone]

Hi,
Yeah the 3d cache thing seens to be the bomb for most people to care about rather than waiting for ms to use crappy e threads efficiently lol

Personally I tend to turn off all things ms to make sure they don't interfere.

 

[Tek-Check]

Meteor Lake chips might not be ready until Week 52 due to EUV scaling challenges. Here is in-depth report explaining it.

Embracing Chaos: The Imperfect Art of Semiconductor Manufacturing And Lithography

5 Longs, Intel EUV Delayed Again, And More

www.semianalysis.com

 

[AlfaPro1337]

It's amazing how every "ecores bad" comment is here from AMD users who will buy an 6+8 zen 5 and love it

The same kind of idiots that complains:

1. Intel is being greedy by making 2S2P (2 socket, 2 platform) back in 1st gen Core i series
2. Intel making the same 4c processor, and paying more
3. Stagnant microarchitecture
4. Withholding their best product instead of releasing it on day one.
5. Idle power is too high
6. Saving power bills, when they OC the snort out of it.

But when it comes to Evil Su and AMD: they aRe bRinGinG coMpeTitIOn.

Aymdiots are paying US$350 for the latest low-end 6c. Pre-Ryzen, at that price, it is a top end product and increased in core count.

No company is paying you to like them.

 

[chrcoluk]

I can't apologize for the 12900KS's terrible power consumption out of the box, certainly they could have done better if they had a more competitive process (smaller node).

A lot of the issue with Intel is that Alder lake is clocked higher than it should be, simply for the sake of generating a lead or a tie on benchmarks. I found limiting my 12900KS to 5.0 and undervolting it dropped the wattage consumed by 100+ W depending on application, and the temperatures don't much exceed 60C in games. If I ran it out of the box, it would be 110C and 300 watts.

Manufacturers certianly seem to have lost interest in power efficiency now, shipping way too far up the power curve, knowing reviewers concentrate mostly on benchmark performance. Even my older 9900k, I saved a chunk of heat/power by reducing its massive over voltage in its shipped state.

 

[BoboOOZ]

Manufacturers certianly seem to have lost interest in power efficiency now, shipping way too far up the power curve, knowing reviewers concentrate mostly on benchmark performance. Even my older 9900k, I saved a chunk of heat/power by reducing its massive over voltage in its shipped state.

I think the interest is still there, but since the most important for the conclusion of the review is the maximum performance, AMD has started playing the juicing game too. Still, good reviewers focus on efficiency and I think AMD are conscious that part of their good sales are coming form their efficiency, too. In fact their parts selling the best are by far the more efficient (5600, 5700x, 5800x3d). I'll be happy to buy a 7800x3d and flip it to 65W from the BIOS.

 

[THU31]

It's amazing how every "ecores bad" comment is here from AMD users who will buy an 6+8 zen 5 and love it

I've used Intel CPUs exclusively for over a decade (and many times before that) and I find E-cores to be completely 100% pointless on desktops.

There is NO benefit. No significant power saving (which is only relevant for battery life anyway), because you'd have to use C-states and core parking, and all those features add latency and cause stutters (even in something as simple as video playback).
Background tasks can run on extra P-cores with no performance penalty in the foreground application. And you could have 16 P-cores with lower power consumption than the 8+16 config.

But I don't have a problem with E-cores existing. I have a problem with Intel shoving them down our throats. Just like they did iGPUs a decade ago. It took so many years for F SKUs to become a thing. Maybe we'll get SKUs without E-cores one day (on the higher end).
Gaming performance is all I care about (productivity in a home environment is completely irrelevant). And for that my preffered config at the moment is 8 P-cores with no E-cores and no HT (for compatibility). 6C/12T is enough, but HT does cause issues occasionaly. Add to that a fixed clockspeed with no power-saving features (I manually switch power plans when needed). Any kind of dynamic switching is disastrous, and that goes for both CPUs and GPUs.

 

[phanbuey]

I've used Intel CPUs exclusively for over a decade (and many times before that) and I find E-cores to be completely 100% pointless on desktops.

There is NO benefit. No significant power saving (which is only relevant for battery life anyway), because you'd have to use C-states and core parking, and all those features add latency and cause stutters (even in something as simple as video playback).
Background tasks can run on extra P-cores with no performance penalty in the foreground application. And you could have 16 P-cores with lower power consumption than the 8+16 config.

But I don't have a problem with E-cores existing. I have a problem with Intel shoving them down our throats. Just like they did iGPUs a decade ago. It took so many years for F SKUs to become a thing. Maybe we'll get SKUs without E-cores one day (on the higher end).
Gaming performance is all I care about (productivity in a home environment is completely irrelevant). And for that my preffered config at the moment is 8 P-cores with no E-cores and no HT (for compatibility). 6C/12T is enough, but HT does cause issues occasionaly. Add to that a fixed clockspeed with no power-saving features (I manually switch power plans when needed). Any kind of dynamic switching is disastrous, and that goes for both CPUs and GPUs.

The 9700K builds that I had (8 non-ht threaded cores) had pretty bad frame pacing issues vs the 8700K -- which had 2 less cores but had HT. I tested this firsthand. Also turning off HT on modern Intel CPUs (7820x, 8700k, 10850K) induces stuttering in games that otherwise run smooth - (Far Cry 5, Borderlands 2, 3). There are a few reviews that show this clearly using frame time graphs (some speculation it has to do with cache prefetching). But disabling HT on recent cpus gives you a performance regression in all but a few outlier games, so not sure why you would want to turn that off.

Also Intel could not be competitive on a 10NM node in mt vs a 5nm Zen 4 if not for e cores -- they simply wouldn't have a competitive product. E produce way less heat, and are 1/4 of the size of P cores, and it leads them down the path of disaggregated design and heterogeneous cores -- which is also a strategic choice. So 10P core-only chip not only would be extremely expensive and uncompetitive in the consumer desktop, but also wouldn't further any innovation.

If you want non HT P-cores only - there's an option to pay the $850 for the 10core/20 thread of sapphire rapids, and then turn off HT -- but then you will be sad when your non-ht very expensive and clock-limited system gets absolutely spanked in literally everything by a 8P/16e hybrid at 40% less power and 40% less cost. It doesn't really make sense IMO.

I see where you're coming from but run a Comet / Alder lake with C states on vs off and test the latency penalty for yourself -- I could find no difference in performance or latency. The implementation has changed - same goes for HT.

 

[THU31]

I play in 4K60 with Vsync and RTSS limiter, so as long as a game has no hitches (and many do, unfortunately), I get a perfectly smooth and consistent experience.

C-states have always been a problem for me (frame skips during video playback, which get much worse with a variable clock speed). Maybe it's different with newer CPUs, or maybe people just don't see it (many people watch 24 FPS movies in 60 Hz and don't see judder).

I don't need 10 cores. I like having 8 right now. I'd absolutely buy a 13700KFC without E-cores if it was $50 cheaper than the KF. Paying for something I'm going to disable sucks.

 

[Der Lokator]

Honestly the Raptor Lake Refresh is welcome, if you can actually push higher max OC clock speeds.
Meteor Lake S: it´s good that they drop it to keep some market share, but 8-core users will skip it.
A lot of people will also skip the 8 core cpu after Meteor Lake S, because it could be the last supported CPU on that platform, so they will just skip the whole platform.

 

[regs]

E-cores are pretty much useless in my use cases.
I would treat this as a 6 core CPU
Now we are back to 2017 where 8700k just came out ?

Those Crestmont cores are quite more powerful than Skylake cores in 8700K. They won't have hyper threading, but 2 real cores are more powerful than a virtual thread. So just E-core block alone would be better than 8700K.