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

[Tigerick]

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 U	Intel Wildcat Lake 15W?	Intel Lunar Lake		Intel Panther Lake 4+0+4
Launch Date		Q1-2024	Q2-2026	Q3-2024		Q1-2026
Model		Intel 150U	Intel Core 7	Core Ultra 7 268V		Core Ultra 7 365
Dies		2	2	2		3
Node		Intel 7 + ?	Intel 18-A + TSMC N6	TSMC N3B + N6		Intel 18-A + Intel 3 + TSMC N6
CPU		2 P-core + 8 E-cores	2 P-core + 4 LP E-cores	4 P-core + 4 LP E-cores		4 P-core + 4 LP E-cores
Threads		12	6	8		8
Max Clock		5.4 GHz	?	5 GHz		4.8 GHz
L3 Cache		12 MB		12 MB		12 MB
TDP		15 - 55 W	15 W ?	17 - 37 W		25 - 55 W
Memory		128-bit LPDDR5-5200	64-bit LPDDR5	128-bit LPDDR5x-8533		128-bit LPDDR5x-7467
Size		96 GB		32 GB		128 GB
Bandwidth				136 GB/s
GPU		Intel Graphics	Intel Graphics	Arc 140V		Intel Graphics
RT		No	No	YES		YES
EU / Xe		96 EU	2 Xe	8 Xe		4 Xe
Max Clock		1.3 GHz	?	2 GHz		2.5 GHz
NPU		GNA 3.0	18 TOPS	48 TOPS		49 TOPS

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.

 

[Hulk]

I’m referring more about the LPE cores. They’re significantly more performant per watt.
The E cores have to get quite a bit more juice just to perform at the same level or slightly better. Which is odd because the LPE cores are just slightly down-clocked Darkmonts. Yet they have a very different profile.

Oh okay. Yes, I noticed that as well.
I can't explain it. Why would Darkmont LPE be significantly more power efficient than Darkmont E?

 

[Hulk]

Cause they are not on Ring so they have different power profile.

I'm not following? What difference specifically make a Darkmont LPE core use about 1 watt less than a Darkmont E core at the same work rate (performance)? A difference like that would seem to indicate a different architecture, node, or both? At first I was thinking super low leakage silicon that doesn't clock high but is very power efficient?

Thing is the LPE's are on the N6 process platform controller tile, which you would think would be less power efficient?

I'm all confused now.

 

[511]

I'm not following? What difference specifically make a Darkmont LPE core use about 1 watt less than a Darkmont E core at the same work rate (performance)? A difference like that would seem to indicate a different architecture, node, or both? At first I was thinking super low leakage silicon that doesn't clock high but is very power efficient?

Thing is the LPE's are on the N6 process platform controller tile, which you would think would be less power efficient?

I'm all confused now.

So here is the brief on MTL/ARL the LP-E is crestmont and are on N6 in SoC without any type of cache to help them.
In LNL/PTL they are in the compute tile on N3B/18A and has special cache Called Memory Side Cache that's for LP-E so they don't starve it's a power optimized cache just to keep power low.

The Darkmont/Skymont are exactly the same cores otherwise like physically the same I have actually matched their dishots.

 

[hemedans]

View attachment 138429
There is this as well

Wouldn't this make PTL-LP Ecore more efficient than cortex A720?

 

[511]

Wouldn't this make PTL-LP Ecore more efficient than cortex A720?

I don't remember A720 SIR2K17 Graph

 

[poke01]

XPS as a sub-brand definitely has one of the best track records out there, with well built machines, well-tuned hardware and firmware and offered with enterprise-grade warranties (sometimes at no extra cost)

The last few XPS models have been questionable. Removed the headphone jack, no viable trackpad and capacitive function row and only usb-c with no other ports

They fixed most of these in the 2026 XPS but still offer only usb-C.

 

[hemedans]

I don't remember A720 SIR2K17 Graph

This is not comparable? Or different subtest?

 

[poke01]

This is not comparable? Or different subtest?
View attachment 138437

Not comparable

 

[511]

This is not comparable? Or different subtest?
View attachment 138437

Different compiler so no

 

[LightningZ71]

Aren't the PTL E cores made on a "fin flex" capable node now? Wouldn't it make sense to flex them to a more efficient configuration with respect to power at the expense of performance?

 

[511]

Aren't the PTL E cores made on a "fin flex" capable node now? Wouldn't it make sense to flex them to a more efficient configuration with respect to power at the expense of performance?

18A doesn't have FinFlex/NanoFlex it's a TSMC Thingy

 

[Magio]

The last few XPS models have been questionable. Removed the headphone jack, no viable trackpad and capacitive function row and only usb-c with no other ports

They fixed most of these in the 2026 XPS but still offer only usb-C.

Don't get me wrong, I loathe the previous couple XPS iterations, the capacitive function row was on its own was just disqualifying for me. Those boneheaded decisions led me to ditching XPS to go with a Thinkpad X9. But while I disliked those models, they were still well built, well tuned and all that jazz (from what I've heard at least). In the same way, I don't dock points from all Thinkpads until the X9 for their trackpoint and odd trackpads buttons, I personally hate that but it's a debatable design choice rather than an objective flaw.

Anyway, to get back on topic the main thing about XPS is people need to stop bringing up their pricing as if it represents PTL's.

 

[Hulk]

So here is the brief on MTL/ARL the LP-E is crestmont and are on N6 in SoC without any type of cache to help them.
In LNL/PTL they are in the compute tile on N3B/18A and has special cache Called Memory Side Cache that's for LP-E so they don't starve it's a power optimized cache just to keep power low.

The Darkmont/Skymont are exactly the same cores otherwise like physically the same I have actually matched their dishots.

This is great thanks.
A couple follow up questions if you don't mind?

First, I assume the Vcore VRM telemetry is a measurement of the voltage regulator for the entire core rail and this additional power is included in the E's per core power, thus making it higher than the LPE's? That would make sense.

Second, can Intel optimize the fab cell structure of the LPE's compared to the E's fab process to tune them for lower power operation? Meaning higher threshold voltage, less leaky transistors, easier to "flip on and off" but result in less max frequency?

If this is what is going on then it is very clever. Intel has essentially built a super low power processor within the low power processor.

The LPE island is NOT simply taking some E's and lowering the frequency. It really is a self-sufficient "island of compute."

 

[511]

This is great thanks.
A couple follow up questions if you don't mind?

First, I assume the Vcore VRM telemetry is a measurement of the voltage regulator for the entire core rail and this additional power is included in the E's per core power, thus making it higher than the LPE's? That would make sense.

Yes

Second, can Intel optimize the fab cell structure of the LPE's compared to the E's fab process to tune them for lower power operation? Meaning higher threshold voltage, less leaky transistors, easier to "flip on and off" but result in less max frequency?

They can but they haven't done it from the die shots I have seen for Lunar Lake it's exactly the same cores aka design someone like Tech Insight can know for sure with teardowns.

If this is what is going on then it is very clever. Intel has essentially built a super low power processor within the low power processor.

The LPE island is NOT simply taking some E's and lowering the frequency. It really is a self-sufficient "island of compute."

Well yeah it's self sufficient compute island.

 

[LightningZ71]

18A doesn't have FinFlex/NanoFlex it's a TSMC Thingy

Intel refers to their RibbonFet tech as a competing solution that supposedly allows variable ribbon configurations on the same die, thick when higher current is needed and thin when better power efficiency is needed. RibbonFet is a feature of 18a.

 

[MerryCherry]

Intel refers to their RibbonFet tech as a competing solution that supposedly allows variable ribbon configurations on the same die, thick when higher current is needed and thin when better power efficiency is needed. RibbonFet is a feature of 18a.

RibbonFET is Intel's name for GAAFET, no?

TSMC styles their GAAFET transistors as Nanosheet.

 

[511]

Intel refers to their RibbonFet tech as a competing solution that supposedly allows variable ribbon configurations on the same die, thick when higher current is needed and thin when better power efficiency is needed. RibbonFet is a feature of 18a.

That's not part of 18A but 14A

 

[Win2012R2]

That's not part of 18A but 14A

You mean specifically variable ribbon config, ie "Turbo Cells"?

 

[511]

You mean specifically variable ribbon config, ie "Turbo Cells"?

Yeah that one

 

[Win2012R2]

Yeah that one

So N2 got a bit of a leg up then with their nanoFlex, which should be V2 now

 

[511]

So N2 got a bit of a leg up then with their nanoFlex, which should be V2 now

Well yeah

 

[coercitiv]

Good luck boys and girls, we're gonna' need it: 256GB base storage on 2026 premium laptops from Samsung.

https://videocardz.com/newz/samsung-galaxy-book6-panther-lake-laptops-priced-for-europe-entry-model-starts-at-e1149-and-book6-ultra-reaches-e3399

 

[511]

Good luck boys and girls, we're gonna' need it: 256GB base storage on 2026 premium laptops from Samsung.

https://videocardz.com/newz/samsung-galaxy-book6-panther-lake-laptops-priced-for-europe-entry-model-starts-at-e1149-and-book6-ultra-reaches-e3399

The cartel guy is price gouging

 

[Magio]

Good luck boys and girls, we're gonna' need it: 256GB base storage on 2026 premium laptops from Samsung.

https://videocardz.com/newz/samsung-galaxy-book6-panther-lake-laptops-priced-for-europe-entry-model-starts-at-e1149-and-book6-ultra-reaches-e3399

We're a couple months away from OEMs bringing back laptops with 4GB of RAM at this rate.

 

[511]

We're a couple months away from OEMs bringing back laptops with 4GB of RAM at this rate.

32 bit windows and 4 GB Ram.