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

Page 946 - 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
939
850
106
Wildcat Lake (WCL) Specs

Intel Wildcat Lake (WCL) is upcoming mobile SoC replacing Raptor Lake-U. WCL consists of 2 tiles: compute tile and PCD tile. It is true single die consists of CPU, GPU and NPU that is fabbed by 18-A process. Last time I checked, PCD tile is fabbed by TSMC N6 process. They are connected through UCIe, not D2D; a first from Intel. Expecting launching in Q1 2026.

Intel Raptor Lake UIntel Wildcat Lake 15W?Intel Lunar LakeIntel Panther Lake 4+4+4
Launch DateQ1-2024Q2-2026Q3-2024Q1-2026
ModelIntel 150UIntel Core 7Core Ultra 7 268VCore Ultra 7 365
Dies2223
NodeIntel 7 + ?Intel 18-A + TSMC N6TSMC N3B + N6Intel 18-A + Intel 3 + TSMC N6
CPU2 P-core + 8 E-cores2 P-core + 4 LP E-cores4 P-core + 4 LP E-cores4 P-core + 4 LP E-cores
Threads12688
Max Clock5.4 GHz?5 GHz4.8 GHz
L3 Cache12 MB12 MB12 MB
TDP15 - 55 W15 W ?17 - 37 W25 - 55 W
Memory128-bit LPDDR5-520064-bit LPDDR5128-bit LPDDR5x-8533128-bit LPDDR5x-7467
Size96 GB32 GB128 GB
Bandwidth136 GB/s
GPUIntel GraphicsIntel GraphicsArc 140VIntel Graphics
RTNoNoYESYES
EU / Xe96 EU2 Xe8 Xe4 Xe
Max Clock1.3 GHz?2 GHz2.5 GHz
NPUGNA 3.018 TOPS48 TOPS49 TOPS






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.



LNL-MX.png
 

Attachments

  • PantherLake.png
    PantherLake.png
    283.5 KB · Views: 24,041
  • LNL.png
    LNL.png
    881.8 KB · Views: 25,529
  • INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg
    INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg
    181.4 KB · Views: 72,437
  • Clockspeed.png
    Clockspeed.png
    611.8 KB · Views: 72,324
Last edited:

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
How is XeSS quality vs DLSS/FSR3 Balanced in terms of image quality?
XeSS XMX is like DLSS3 or a bit better for the same resolution scaling but worse than FSR4/DLSS 4 no info on framegen.

Right now sure, but it's something they need to fix, because it's future debt.
well Intel/AMD needs to ship fat cores not skinny cores in terms of resourcing and it will blow the area budget if you have to cater both AVX-512/Int performance
 

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
Still well behind RDNA4 currently, and no, there will only be Xe3P by the time GFX13 is out.
Medusa Premium will show just compromised AMD iGPUs have been, reality comes for us all eventually.
If only you looked at Xe3P patches that Xe3P is the largest upgrade since Xe from the limited look i had the patches it's basically 2X in terms of throughput in RT/XMX/FP32
MDSP obliterates PTL in CPU/GPU perf, not even remotely fair.
I'd wait 2 years for that personally given the slowing iteration rate in laptops.
MDSP is a Big Fat Bus so not a fair comparison like you said Comparison RZL-AX should be the proper comparison
Xe4 is Celestial right? The nice thing for Intel is that game performance is not so much of a moving target as resolution has topped out at 4k and will be there for a long while. Plus, some gaming is doing at a lower computational level and AI scaled up. While the game engines are getting more compute heavy they are coded with the "average" system in mind.

My point is that Alchemist and Battlemage are basically playing the same games but Battlemage provides much more performance for the money. My brother is happily gaming on a B570. If there aren't obvious stutters he doesn't care. He doesn't even care as long as it doesn't happen to much. There is still a market for the high end cards but the lower end card are creeping up on them.
It's not celestial Celestial is Xe3P Xe4 is well a 2028ish IP
 

fastandfurious6

Senior member
Jun 1, 2024
895
1,033
96
Some real benchmarks from Digital Foundry. Panther Lake iGPU looking pretty impressive.

very weird delivery format. they keep cutting the live video FPS every 5 seconds to interject their faces. no static slides. only 2 games tested with Halo, then replaced by RTX3050? lol


evidently Intel is running a carefully crafted PR campaign at the moment.... with another 2 weeks until review embargo lift (two embargos lol)


TBH best thing AMD should do right now:

direct leak MDSP benchmark
 

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
very weird delivery format. they keep cutting the live video FPS every 5 seconds to interject their faces. no static slides. only 2 games tested with Halo, then replaced by RTX3050? lol


evidently Intel is running a carefully crafted PR campaign at the moment.... with another 2 weeks until review embargo lift (two embargos lol)


TBH best thing AMD should do right now:

direct leak MDSP benchmark
There is not even a silicon Yet they will be sharing projections lmao
 

ToTTenTranz

Senior member
Feb 4, 2021
893
1,490
136
Idek what does that even mean
In this case it's reverse Osborne. AMD keeps announcing new products whose iGPU they state is on life support (no FSR4, no Redstone).


It's pretty disgraceful and honestly they deserve every sale lost from the general public's perception that their feature and software support is now terrible.



I can't see anything in that page about scaling gaming performance above 35W.
And for example Strix Halo can scale nT CPU performance all the way up to 115W, especially the 16 core 395 version. OTOH gaming performance won't really go up above 60W because then it's the caches and RAM bandwidth that can't keep up with GPU throughput.


How is XeSS quality vs DLSS/FSR3 Balanced in terms of image quality?

Edit: XeSS version update they did a while ago changed the meaning of the naming. XeSS quality post-update is lower than pre-update. Balanced FSR may be comparable to XeSS Quality.
AFAIK they didn't lower the IQ for the same setting, they lowered the base resolution for each setting with the conviction that they're preserving IQ.




1768403062487.png

The problem is we now have FSR3 Quality and DLSS4 Quality with a higher base resolution than XeSS3 Quality. The reverse problem is that we now have DLSS Quality "M" that looks better than DLSS Quality "K", both with distinct performance results, and I honestly don't know which situation is worst.


It's 128-bit LPDDR5X-9600
So no 192bit LPDDR6 for Medusa Premium / Halo?
 
  • Like
Reactions: Elfear

OneEng2

Senior member
Sep 19, 2022
971
1,183
106
So ....

After reading through lots of posts here, and others in different forums, and looking at the information put out at CES about PTL, seems like Intel is betting on gaming laptop market with PTL?

So, is PTL a well position product? From my understanding, it doesn't compete with a dedicated graphics card solution in gaming. It is a pretty big piece of silicon (CPU tile on 18A at ~114mm^2 compared to Zen 5 at 70.6mm^2) with a bunch of other tiles and packaging. Seems to be a pretty pricy solution (from a cost perspective).

Intel went all-in on the graphics update from what I can see. From a customer standpoint, are graphics really that important in the general thin and light category .... or even the general laptop market?

Just curios on peoples thoughts on the positioning of this product?
 

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
So ....

After reading through lots of posts here, and others in different forums, and looking at the information put out at CES about PTL, seems like Intel is betting on gaming laptop market with PTL?

So, is PTL a well position product? From my understanding, it doesn't compete with a dedicated graphics card solution in gaming. It is a pretty big piece of silicon (CPU tile on 18A at ~114mm^2 compared to Zen 5 at 70.6mm^2) with a bunch of other tiles and packaging. Seems to be a pretty pricy solution (from a cost perspective).

Intel went all-in on the graphics update from what I can see. From a customer standpoint, are graphics really that important in the general thin and light category .... or even the general laptop market?

Just curios on peoples thoughts on the positioning of this product?
the comparisons are kind of wrong Strix Point for example is 230ish mm2 N4P Panther Lake is 114mm2 18A+(54 N3E/33 I3)mm2 GPU silicon and and 48mm2 N6 PCD plus advanced packing roughly 218mm2 of silicon everything else is filler or base die included in advanced packing cost
 
  • Like
Reactions: DKR

desrever

Senior member
Nov 6, 2021
335
820
136
So ....

After reading through lots of posts here, and others in different forums, and looking at the information put out at CES about PTL, seems like Intel is betting on gaming laptop market with PTL?

So, is PTL a well position product? From my understanding, it doesn't compete with a dedicated graphics card solution in gaming. It is a pretty big piece of silicon (CPU tile on 18A at ~114mm^2 compared to Zen 5 at 70.6mm^2) with a bunch of other tiles and packaging. Seems to be a pretty pricy solution (from a cost perspective).

Intel went all-in on the graphics update from what I can see. From a customer standpoint, are graphics really that important in the general thin and light category .... or even the general laptop market?

Just curios on peoples thoughts on the positioning of this product?
My thinking right now is the current performance numbers coming out are all affiliated with Intel marketing. Real average across a lot of games will be lower. But even ignoring that, there is some key problems for the platform that has nothing to do with the performance of the silicon:
  • Windows 11 isn't driving sales and AI PCs are just not wanted by consumers.
  • Cost of the entire laptop is going to be much more expensive due to not only the cost of PTL itself but all the inflated pricing of everything else.
  • There is still a lot of consumer inventory of previous gen laptops with previous gen MSRPs before ram cost exploded
  • IGPU performance has never been the driver of sales growth, AMD lead this for years and gained nothing, everyone wants Nvidia if they want to game.
  • Battery life will still be below ARM competition for people who really care.
 

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
My thinking right now is the current performance numbers coming out are all affiliated with Intel marketing. Real average across a lot of games will be lower. But even ignoring that, there is some key problems for the platform that has nothing to do with the performance of the silicon:
  • Windows 11 isn't driving sales and AI PCs are just not wanted by consumers.
Yeah no one like Win11
  • Cost of the entire laptop is going to be much more expensive due to not only the cost of PTL itself but all the inflated pricing of everything else.
Ram and Storage balloning up
  • There is still a lot of consumer inventory of previous gen laptops with previous gen MSRPs before ram cost exploded
  • IGPU performance has never been the driver of sales growth, AMD lead this for years and gained nothing, everyone wants Nvidia if they want to game.
it's cost and always has been
  • Battery life will still be below ARM competition for people who really care.
Yeah no that's not true under load there is no ARM no x86 battery just dies in a hour for mixed usage LNL is literally ARM level battery life
 

Khato

Golden Member
Jul 15, 2001
1,350
436
136
The 114mm^2 PTL 'CPU tile' is actually the SoC tile. PTL is kind of a hybrid between the LNL and MTL/ARL design. So the PTL SoC tile + graphics tile is close to the LNL compute tile, but contains quite a few things that were in the LNL SoC tile. Either way, contains far, far more logic than the 70.6mm^2 Zen5 CCD.

Despite claims to the contrary, PTL is a relatively 'cheap' mainstream platform. Will be interesting to see how popular the 12Xe parts are as the incremental cost to Intel is far less than sale price.

One point of interest regarding current inflated prices on memory - it just makes low end discrete GPUs that much more expensive compared to integrated. A low end gaming laptop with discrete is 16GB system + 8GB GPU, so 50% more memory than integrated.
 
  • Like
Reactions: DKR

poke01

Diamond Member
Mar 8, 2022
4,736
6,075
106
Yeah no that's not true under load there is no ARM no x86 battery just dies in a hour for mixed usage LNL is literally ARM level battery life
LNL is only good in regards to battery life if you get the 226V and if the screen is 1080p or 1200p lcd.
 

poke01

Diamond Member
Mar 8, 2022
4,736
6,075
106
That's absolutely not true for only SKU lol also picking OLED for battery life is not a good decision obvious reasons.
IMG_3111.jpeg

I mean you can see here. Add a high resolution panel ( none of those are OLED) or pick a non Ultra 5 SKU and the battery LNL reduces a lot.

The Acer is standard 60hz FHD.
 

DavidC1

Platinum Member
Dec 29, 2023
2,093
3,215
106
Intel costs are not directly compatible with anyone else because filling their fabs is more important than little BoM differences. Even if it costs them 20% more on 18A, that's still utilization versus if they used TSMC and isn't being utilized so the loss is 100% for them. This is the same logic behind why lowering costs to keep volume share is a good strategy(because not sold chip = 0%), or why perceived added costs of an iGPU on a die more than counters it because not having it means you might lose an entire SoC design win. You lose 5% versus you lose 100%.

Pat's logic behind using 3rd party fabs was to hopefully motivate internal process teams to do better because they aren't fed by inefficiencies due to no competition. But eventually everything that can come to Intel should come to Intel.
LNL is only good in regards to battery life if you get the 226V and if the screen is 1080p or 1200p lcd.
There's no difference between 226V and 258V in this regard though. The real impact is the screen choice.
Battery life will still be below ARM competition for people who really care.
The ARM Windows market doesn't matter. It's Windows x86 vs Apple ARM.

Because you see, if you are going to deal with Windows, you might as well get the good side as well, which is worry-free compatibility. Pre-Lunarlake that argument held more merit, but Lunar and Panther makes that irrelevant. And if you are ok with incompatibility, you might as well get Apple. Intel/AMD still shouldn't act like Qualcomm or standard ARM vendors don't exist, because eventually a prolonged deficit will result in inevitable decline for them. Once the transition happens, then it's pretty much permanent.

They should act soon when they are in an ok position, not when things are not. It's just like the people in movies where everyone rushes to get out of a disaster prone city, when the ideal thing is to do it BEFORE the ignorant do.
 
Last edited:

Doug S

Diamond Member
Feb 8, 2020
3,789
6,714
136
It’s because Qualcomm is incompetent. No good GPU IP, basically no GPGPU for pro apps and games don't work. You can't just put in a good CPU and neglect everything else.

It isn't just Qualcomm's incompetence, though that isn't helping them. They are at Microsoft's mercy compatibility wise, and at the mercy of developers quality of port wise and support wise.

There just isn't a good reason to swim against the x86 tide and run Windows on ARM. Any performance or power advantage you might get is temporary, and just isn't worth the headaches (both real and potential like "what if Microsoft loses interest in ARM due to low sales, or the next version of your must use app drops ARM support for that reason?")
 

511

Diamond Member
Jul 12, 2024
5,308
4,727
106
It isn't just Qualcomm's incompetence, though that isn't helping them. They are at Microsoft's mercy compatibility wise, and at the mercy of developers quality of port wise and support wise.

There just isn't a good reason to swim against the x86 tide and run Windows on ARM. Any performance or power advantage you might get is temporary, and just isn't worth the headaches (both real and potential like "what if Microsoft loses interest in ARM due to low sales, or the next version of your must use app drops ARM support for that reason?")
Microsoft has more to worry about considering many enterprises are not happy with MSFT
View attachment 136716

I mean you can see here. Add a high resolution panel ( none of those are OLED) or pick a non Ultra 5 SKU and the battery LNL reduces a lot.

The Acer is standard 60hz FHD.
And so are MacBooks 60Ghz 120Hz is going to burn more power for the same display uhh there are various laptops with much better battery than M4 air in this one here are those

A
 

poke01

Diamond Member
Mar 8, 2022
4,736
6,075
106
And so are MacBooks 60Ghz 120Hz is going to burn more power for the same display uhh there are various laptops with much better battery than M4 air in this one here are those

A
all those have 85% less pixels to push than the M4 air. The Dell is pretty good tho since it’s 120hz.
 

poke01

Diamond Member
Mar 8, 2022
4,736
6,075
106
I’m interested to see the Dell XPS 2026 with the 1200p display, hopefully it’s better than the Dell above. Since the new panel goes down to 1Hz