Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads

Page 134 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Tigerick

Senior member
Apr 1, 2022
677
559
106
PPT1.jpg
PPT2.jpg
PPT3.jpg



As Hot Chips 34 starting this week, Intel will unveil technical information of upcoming Meteor Lake (MTL) and Arrow Lake (ARL), new generation platform after Raptor Lake. Both MTL and ARL represent new direction which Intel will move to multiple chiplets and combine as one SoC platform.

MTL also represents new compute tile that based on Intel 4 process which is based on EUV lithography, a first from Intel. Intel expects to ship MTL mobile SoC in 2023.

ARL will come after MTL so Intel should be shipping it in 2024, that is what Intel roadmap is telling us. ARL compute tile will be manufactured by Intel 20A process, a first from Intel to use GAA transistors called RibbonFET.



Comparison of upcoming Intel's U-series CPU: Core Ultra 100U, Lunar Lake and Panther Lake

ModelCode-NameDateTDPNodeTilesMain TileCPULP E-CoreLLCGPUXe-cores
Core Ultra 100UMeteor LakeQ4 202315 - 57 WIntel 4 + N5 + N64tCPU2P + 8E212 MBIntel Graphics4
?Lunar LakeQ4 202417 - 30 WN3B + N62CPU + GPU & IMC4P + 4E08 MBArc8
?Panther LakeQ1 2026 ??Intel 18A + N3E3CPU + MC4P + 8E4?Arc12



Comparison of die size of Each Tile of Meteor Lake, Arrow Lake, Lunar Lake and Panther Lake

Meteor LakeArrow Lake (20A)Arrow Lake (N3B)Arrow Lake Refresh (N3B)Lunar LakePanther Lake
PlatformMobile H/U OnlyDesktop OnlyDesktop & Mobile H&HXDesktop OnlyMobile U OnlyMobile H
Process NodeIntel 4Intel 20ATSMC N3BTSMC N3BTSMC N3BIntel 18A
DateQ4 2023Q1 2025 ?Desktop-Q4-2024
H&HX-Q1-2025
Q4 2025 ?Q4 2024Q1 2026 ?
Full Die6P + 8P6P + 8E ?8P + 16E8P + 32E4P + 4E4P + 8E
LLC24 MB24 MB ?36 MB ??8 MB?
tCPU66.48
tGPU44.45
SoC96.77
IOE44.45
Total252.15



Intel Core Ultra 100 - Meteor Lake

INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg

As mentioned by Tomshardware, TSMC will manufacture the I/O, SoC, and GPU tiles. That means Intel will manufacture only the CPU and Foveros tiles. (Notably, Intel calls the I/O tile an 'I/O Expander,' hence the IOE moniker.)

Clockspeed.png
 

Attachments

  • PantherLake.png
    PantherLake.png
    283.5 KB · Views: 23,968
  • LNL.png
    LNL.png
    881.8 KB · Views: 25,440
Last edited:

Thunder 57

Platinum Member
Aug 19, 2007
2,702
3,873
136
The real exciting thing in MTL is the efficiency, but the idiots at Intel are still keeping it under wraps. Sad.

The longer Intel stays quiet about MTL the more I think it will be a letdown in some area. Even in the interview a few pages ago, she described the core as "holistic". I think it will do well with regards to battery life, but I am not expecting it to do as well with performance.
 

TESKATLIPOKA

Platinum Member
May 1, 2020
2,369
2,863
136
The longer Intel stays quiet about MTL the more I think it will be a letdown in some area. Even in the interview a few pages ago, she described the core as "holistic". I think it will do well with regards to battery life, but I am not expecting it to do as well with performance.
Intel needs to fix efficiency first.
screenshot_2-png.86781

If MTL can significantly increase performance at low TDPs, let's say 65W performance of 13700H at 45W because of higher clocks, then that's pretty good.

Against 4C+8c Strix Point It will certainly lose in MT performance and likely also in efficiency, but we know It takes a long time to see first laptops with AMD's latest chips, so It's not such a big problem.
 
  • Like
Reactions: Tlh97 and Saylick

mikk

Diamond Member
May 15, 2012
4,152
2,164
136
The longer Intel stays quiet about MTL the more I think it will be a letdown in some area. Even in the interview a few pages ago, she described the core as "holistic". I think it will do well with regards to battery life, but I am not expecting it to do as well with performance.


Quiet? They shared basically all technical details at innovation last month including the launch date. Do you expect official benchmarks prior to the launch? Is this your first launch you are watching? Software/firmware probably isn't even final at this point. ST performance is down by the looks of it, maybe 10%, so there will be at least one letdown. MT looks fine on Geekbench.
 

Hulk

Diamond Member
Oct 9, 1999
4,262
2,064
136
Quiet? They shared basically all technical details at innovation last month including the launch date. Do you expect official benchmarks prior to the launch? Is this your first launch you are watching? Software/firmware probably isn't even final at this point. ST performance is down by the looks of it, maybe 10%, so there will be at least one letdown. MT looks fine on Geekbench.
Not quiet but it is kind of unusual for Intel not to provide a couple of performance charts. You know, the ones with unclear labels on the axis that just show trends but it is still something. Intel provided lots of details about everything except performance. They wouldn't even disclose what changes they made to the P and E cores. Architectural day was more of advertising day, providing more "it's gonna be great hype" than actually technical information.

And yes, Intel has shown official benchmarks in the past when they know they have a winner. Remember the Conroe preview?

None of this really means anything though. MTL is a big undertaking for Intel. In my opinion if they can improve on their current mobile product stack it's a win as this product looks to be a good foundation for them moving forward.
 

Thunder 57

Platinum Member
Aug 19, 2007
2,702
3,873
136
Quiet? They shared basically all technical details at innovation last month including the launch date. Do you expect official benchmarks prior to the launch? Is this your first launch you are watching? Software/firmware probably isn't even final at this point. ST performance is down by the looks of it, maybe 10%, so there will be at least one letdown. MT looks fine on Geekbench.

Not quiet but it is kind of unusual for Intel not to provide a couple of performance charts. You know, the ones with unclear labels on the axis that just show trends but it is still something. Intel provided lots of details about everything except performance. They wouldn't even disclose what changes they made to the P and E cores. Architectural day was more of advertising day, providing more "it's gonna be great hype" than actually technical information.

And yes, Intel has shown official benchmarks in the past when they know they have a winner. Remember the Conroe preview?

None of this really means anything though. MTL is a big undertaking for Intel. In my opinion if they can improve on their current mobile product stack it's a win as this product looks to be a good foundation for them moving forward.

The Conroe preview is exactly what I was thinking of. If they had a winner like that, they would be shouting from the rooftops.
 

FangBLade

Member
Apr 13, 2022
199
395
106
If something drastic doesn't change in terms of efficiency (I'm putting my last hopes on MTL, I know it's funny), I think I'll switch to an ARM variant of a laptop. Qualcomm hardware looks impressive according to leaks, and if the software is up to par, I think I could say goodbye to x86 design on a laptop. They consume too much energy.
 

Hulk

Diamond Member
Oct 9, 1999
4,262
2,064
136
If something drastic doesn't change in terms of efficiency (I'm putting my last hopes on MTL, I know it's funny), I think I'll switch to an ARM variant of a laptop. Qualcomm hardware looks impressive according to leaks, and if the software is up to par, I think I could say goodbye to x86 design on a laptop. They consume too much energy.
Isn't the display the largest power draw in a laptop? Also aren't the Zen 4 laptops really efficient?
 

SiliconFly

Golden Member
Mar 10, 2023
1,058
541
96
Isn't the display the largest power draw in a laptop? Also aren't the Zen 4 laptops really efficient?
The display & CPU/GPU are the main components that use lots of power. Next comes RAM, SSD, WiFi, Speakers, etc but they consume significantly less power under full load than the former. Rest of the components not so much.
 

Hulk

Diamond Member
Oct 9, 1999
4,262
2,064
136
When was the last Intel chip where we had good leaks on QS samples "in the wild" before launch?
I thought Golden Cove looks pretty good in the leaks as well as the claims that Gracemont performed about like non-HT Skylake IPC-wise, which I also thought was impressive considering how far Atom had come.

Before that I think Nahalem looked pretty good. Conroe before that. And before that the original Pentium with it being the first superscalar CPU was looking kind of ground breaking. But of course this is completely subject and these are the ones that stick out in my mind.
 

Thibsie

Senior member
Apr 25, 2017
760
819
136
I thought Golden Cove looks pretty good in the leaks as well as the claims that Gracemont performed about like non-HT Skylake IPC-wise, which I also thought was impressive considering how far Atom had come.

Before that I think Nahalem looked pretty good. Conroe before that. And before that the original Pentium with it being the first superscalar CPU was looking kind of ground breaking. But of course this is completely subject and these are the ones that stick out in my mind.
Pentium, the first superscalar cpu ?
Nope. The first complete x86 superscalar cpu, yeah.
 

Hulk

Diamond Member
Oct 9, 1999
4,262
2,064
136
Pentium, the first superscalar cpu ?
Nope. The first complete x86 superscalar cpu, yeah.
You got me. I should have been more specific and wrote that the Pentium was significant in one way because it was the first mainstream computer processor that was superscalar. It was also significant in that it used micro-op fusion in order to better to compete with the Power PC RISC based processors. The micro-ops made the old RISC x86 architecture "RISC-like" when it came to moving instructions more quickly through the pipeline.

I guess now I should also define "mainstream" as indicated in the sentence above. By mainstream I mean a processor that eventually ended up in tens of millions (if not more) of homes, businesses, etc...

I guess now I should also define homes, businesses, etc... By "homes" I mean the places where most people live and businesses are building where people don't generally live but a place of work.

I guess now I should define "live" and "work." By "live" I mean the place that they consider the place they sleep, bathe, eat, relax, etc... By "work" I mean that as most people need to do things for others in order to receive money, the business is the place where they generally go to trade their time and skills for money so that they may pay for their home, food, and other items required for survival.

I guess now I should define what I mean by "survival"....

Obviously I'm kidding here. I'm just highlighting the fact that we need to be very specific when writing in forums!

I do appreciate your comment actually as I'm always up to learn. Just couldn't resist the opportunity to have some fun.
 

mikk

Diamond Member
May 15, 2012
4,152
2,164
136
Not quiet but it is kind of unusual for Intel not to provide a couple of performance charts. You know, the ones with unclear labels on the axis that just show trends but it is still something.

Do you have examples from previous generations? I'm not aware of it. Sure we got leaks sometimes prior to the launch but usually it's the launch day or review reveal day when we get benchmarks or real life figures, SKU specifications, slides with more technical infos and things like that, this is the purpose of a launch. Also it's not so easy for a mobile launch, a lot depends on the implementation. Performance or battery life varies a lot depending on the device and software state. First devices are 8 weeks away.
 

Thunder 57

Platinum Member
Aug 19, 2007
2,702
3,873
136
I don't have any of the old performance claim charts as they are immediately old news once reviews are out but I do have a lot of examples of how Intel used to give us actual information pre-launch. Here are some I have on hand.
View attachment 88031

View attachment 88032
View attachment 88033
View attachment 88034
View attachment 88035
View attachment 88036
View attachment 88037View attachment 88038

Lol at "Intel HD Boost". SSE4.1 sure makes the high def look super duper high def!
 
  • Haha
Reactions: Tlh97 and inf64

mikk

Diamond Member
May 15, 2012
4,152
2,164
136
I don't have any of the old performance claim charts as they are immediately old news once reviews are out but I do have a lot of examples of how Intel used to give us actual information pre-launch. Here are some I have on hand.

There are basically zero changes to the big core and small changes to the E-cores, maybe watch some of the Intel sessions, they are available on Intel.com. It's a tick release when it comes to CPU cores like Ivy Bridge which is missing in the slide. It's the same with Meteor Lake. Haswell or Icelake was a tock, a major change, of course there was more to show in regards to CPU cores. I think you are expecting too much from Meteor Lake. There are no CPU performance figures (beside the IPC claims) from Meteor Lake because it is underwhelming. Intel only mentioned three things all the time: efficiency, GPU and AI.
 

Hulk

Diamond Member
Oct 9, 1999
4,262
2,064
136
There are basically zero changes to the big core and small changes to the E-cores, maybe watch some of the Intel sessions, they are available on Intel.com. It's a tick release when it comes to CPU cores like Ivy Bridge which is missing in the slide. It's the same with Meteor Lake. Haswell or Icelake was a tock, a major change, of course there was more to show in regards to CPU cores. I think you are expecting too much from Meteor Lake. There are no CPU performance figures (beside the IPC claims) from Meteor Lake because it is underwhelming. Intel only mentioned three things all the time: efficiency, GPU and AI.
I'm only expecting them to be available upon release and for them to be more efficient that Raptor parts at ISO performance. I realize the transition to a tiled architecture is a big deal in itself.
 

mikk

Diamond Member
May 15, 2012
4,152
2,164
136
Any thoughts on this?


Arrow Lake-H belongs to C065 which has different (less) ISA features than ARL-S C066. How does it makes sense? Usually it means different CPU architecture. They can't use Redwood Cove/Crestmont for ARL-H on Intel 3 or 20A?
 

Geddagod

Golden Member
Dec 28, 2021
1,159
1,033
106
I don't have any of the old performance claim charts as they are immediately old news once reviews are out but I do have a lot of examples of how Intel used to give us actual information pre-launch. Here are some I have on hand.
View attachment 88031

View attachment 88032
View attachment 88033
View attachment 88034
View attachment 88035
View attachment 88036
View attachment 88037View attachment 88038
Intel stopped giving more technical info as they lost the lead in both foundry and CPU design
 

Kepler_L2

Senior member
Sep 6, 2020
371
1,419
106
Any thoughts on this?


Arrow Lake-H belongs to C065 which has different (less) ISA features than ARL-S C066. How does it makes sense? Usually it means different CPU architecture. They can't use Redwood Cove/Crestmont for ARL-H on Intel 3 or 20A?
I think that's a rebranded MTL refresh on Intel 3.
 
  • Like
Reactions: igor_kavinski
Jan 12, 2021
30
64
61
Any thoughts on this?


Arrow Lake-H belongs to C065 which has different (less) ISA features than ARL-S C066. How does it makes sense? Usually it means different CPU architecture. They can't use Redwood Cove/Crestmont for ARL-H on Intel 3 or 20A?
LPE-core are the same as MTL, ARL-S does not have it, so it does not have this limitation
 

dullard

Elite Member
May 21, 2001
25,111
3,481
126
Any thoughts on this?


Arrow Lake-H belongs to C065 which has different (less) ISA features than ARL-S C066. How does it makes sense? Usually it means different CPU architecture. They can't use Redwood Cove/Crestmont for ARL-H on Intel 3 or 20A?
Tom's Hardware had a small write up on it yesterday. Seems like a minor difference, speculated due to ultra-low power SOC processors.

Seems like the affected instructions include hardware security (formats used mostly in China), some AI, and debugging/optimization when programming software. For those of us outside China the main difference will be that the mobile version will be slower than desktop--if those AI instructions are used.

That said, I can't stand fragmentations like that. If software developers write for the lowest common denominator, then no processor gets the improvements.
 
  • Like
Reactions: igor_kavinski