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

[Tigerick]

Wildcat Lake (WCL) Preliminary Specs

Intel Wildcat Lake (WCL) is upcoming mobile SoC replacing ADL-N. 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 Q2/Computex 2026. In case people don't remember AlderLake-N, I have created a table below to compare the detail specs of ADL-N and WCL. Just for fun, I am throwing LNL and upcoming Mediatek D9500 SoC.

		Intel Alder Lake - N	Intel Wildcat Lake	Intel Lunar Lake		Mediatek D9500
Launch Date		Q1-2023	Q2-2026 ?	Q3-2024		Q3-2025
Model		Intel N300	?	Core Ultra 7 268V		Dimensity 9500 5G
Dies		2	2	2		1
Node		Intel 7 + ?	Intel 18-A + TSMC N6	TSMC N3B + N6		TSMC N3P
CPU		8 E-cores	2 P-core + 4 LP E-cores	4 P-core + 4 LP E-cores		C1 1+3+4
Threads		8	6	8		8
Max Clock		3.8 GHz	?	5 GHz
L3 Cache		6 MB	?	12 MB
TDP		7 W	Fanless ?	17 W		Fanless
Memory		64-bit LPDDR5-4800	64-bit LPDDR5-6800 ?	128-bit LPDDR5X-8533		64-bit LPDDR5X-10667
Size		16 GB	?	32 GB		24 GB ?
Bandwidth			~ 55 GB/s	136 GB/s		85.6 GB/s
GPU		UHD Graphics		Arc 140V		G1 Ultra
EU / Xe		32 EU	2 Xe	8 Xe		12
Max Clock		1.25 GHz		2 GHz
NPU		NA	18 TOPS	48 TOPS		100 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.

 

[Magio]

Didn't N3B density also get nerfed in favor of yields in actual products?

 

[adroc_thurston]

Didn't N3B density also get nerfed in favor of yields in actual products?

No.
It's just far off from what they've promised in like 2019.

 

[511]

Worst how (other than cost as discussed)?

It's the least performant N3 Node out of the family

 

[mikegg]

Have you tried LNL or are you still stuck at RPL

Yes I have. Throttle like crazy.

I have both and I won't use Mac Cause for personal use ever I can't use half of my stuff and tools.

What stuff and tools? Take a picture of your Macbook running Anandtech and this comment.

Yeah that's not true only reason you get MacBook experience is Apple actually integrating stuff correctly which Intel/AMD can't do and neither Qcom

Integrating what stuff correctly? It's the CPU and SoC efficiency. All that matters for a good mass market laptop experience. Qualcomm will definitely do it well as soon as Elite 2 come out.

 

[511]

What stuff and tools? Take a picture of your Macbook running Anandtech and this comment.

My work laptop can't do that

 

[511]

Integrating what stuff correctly? It's the CPU and SoC efficiency. All that matters for a good mass market laptop experience. Qualcomm will definitely do it well as soon as Elite 2 come out.

It's more than SoC there are other components like Power Delivery(PMIC/VRM) / Quality Control / choice of materials/ PCB Suppliers etc as well

 

[poke01]

It's more than SoC there are other components like Power Delivery(PMIC/VRM) / Quality Control / choice of materials/ PCB Suppliers etc as well

isn't that what Intel Evo is?

 

[511]

isn't that what Intel Evo is?

That program is still on? I think that was one way to do it

 

[DavidC1]

It's more than SoC there are other components like Power Delivery(PMIC/VRM) / Quality Control / choice of materials/ PCB Suppliers etc as well

SoC does play a factor. You do need the other parts optimized too, but the CPU acts as the owner of the "leash" so the rest of the platform stays at lower power. Intel is significantly better than AMD at this, partly because they have the experience of entering the mobile market. ARM is better because either they do, or they get kicked out of the market. Because in terms of parts the manufacturer has same choice whether ARM or x86. Look at the power management chips. They are all commonly available parts.

Simply AMD/Intel has no motivation to do so. Actually I argue they do, but no immediate threat exists in the short term. What would wake them up is if the corrupt and fat "Just Us" system decides one day x86 licenses are all void and null and anyone can make x86 CPUs.

Qualcomm ain’t getting any meaningful market share with those pathetic iGPUs.

Qualcomm isn't getting it because compatibility issues still exist. When you buy a $1000 laptop and it doesn't work for you, and you find out the reason, you'll swear off the product. Everything else, like the technical specs are made irrelevant. 1000x x 0 is zero. And Lunarlake gets close enough that they are made irrelevant. Intel made them irrelevant with Bay Trail during Windows 8 era, they are doing it again. Only problem with Intel is they address it and they give it up, looking it from greenback colored glasses rather than thinking "Oh hey, maybe this will result in a boon for the market in general, let's keep making them". And AMD, despite all their strength, ignores the low power market entirely. Both vendors are retarded IMO.

We'll see whether Zen 6 actually does better than Novalake, because I don't think it will. It'll likely be no better than current Arrow vs Zen 5. Way different approaches, but still end up mostly similar.

 

[511]

SoC does play a factor. You do need the other parts optimized too, but the CPU acts as the owner of the "leash" so the rest of the platform stays at lower power. Intel is significantly better than AMD at this, partly because they have the experience of entering the mobile market. ARM is better because either they do, or they get kicked out of the market. Because in terms of parts the manufacturer has same choice whether ARM or x86. Look at the power management chips. They are all commonly available parts.

Laptops vendor don't like PMIC nor memory on package LNL is a one off cause Intel is selling ram at the price it bought and the PMIC are expensive compared to regular VRM and QCOM is pushing both With X2 Elite.

 

[mikegg]

It's more than SoC there are other components like Power Delivery(PMIC/VRM) / Quality Control / choice of materials/ PCB Suppliers etc as well

Yea... no. Windows laptops routinely achieve Macbook like idle power. It's when they actually need to compute (SoC, CPU), AMD and Intel fall apart.

 

[511]

Yea... no. Windows laptops routinely achieve Macbook like idle power. It's when they actually need to compute (SoC, CPU), AMD and Intel fall apart.

 

[DavidC1]

Laptops vendor don't like PMIC nor memory on package LNL is a one off cause Intel is selling ram at the price it bought and the PMIC are expensive compared to regular VRM and QCOM is pushing both With X2 Elite.

Like @mikegg says, it's not so good when it's on load. Actually Icelake had very decent idle, it just couldn't get good battery life even with the slightest bit of load, so what's the point of low idle?

Although I'm not totally pessimistic as he is in this regard. Lunarlake is close enough in the battery department to not go ARM for laptops. Also some of the deficiencies of x86 is the other counterpart, which is Windows. Last Windows Microsoft tried to make decent was Windows 7. After that it was a hodpodge mess of inconsistent UI directions and starting in Windows 10 it was in-your face spyware and adware.

 

[DavidC1]

15% for Novalake vs Panther would be acceptable if that comes at the same clockspeeds, or even lower. Panther is already few % above Arrow per clock.

 

[adroc_thurston]

15% for Novalake vs Panther would be acceptable if that comes at the same clockspeeds, or even lower

Why the hell would it be acceptable when the rest of the world is huffing juice and going fast

 

[Meteor Late]

Why the hell would it be acceptable when the rest of the world is huffing juice and going fast

Yeah it's hilarious, isn't it? everyone else is getting more clock speed along with IPC increases, whether they are smaller or larger, but somehow it is acceptable for Intel to be able to maintain clock speed and not increase it.

 

[511]

Yeah it's hilarious, isn't it? everyone else is getting more clock speed along with IPC increases, whether they are smaller or larger, but somehow it is acceptable for Intel to be able to maintain clock speed and not increase it.

Depending on the Intel core tbh Skymont got 2 gens of IPC Improvement last year but everyone forgets that Lion Cove barely got 10%

 

[Magio]

Depending on the Intel core tbh Skymont got 2 gens of IPC Improvement last year but everyone forgets that Lion Cove barely got 10%

P-core has been having a tough couple years for sure. If Coyote Cove isn't a return to form Intel is just going to stink it up in ST up until Unified Core in 2028+.

 

[Geddagod]

P-core has been having a tough couple years for sure. If Coyote Cove isn't a return to form Intel is just going to stink it up in ST up until Unified Core in 2028+.

No guarantee unified core will be all that great either tbh. Esp in perf, even if we see a large uplift for Intel in perf/mm2 and perf/watt.
I don't think skymont is the best core in it's area class either. And perf/area is by far Skymont's best attribute.

 

[Magio]

No guarantee unified core will be all that great either tbh. Esp in perf, even if we see a large uplift for Intel in perf/mm2 and perf/watt.
I don't think skymont is the best core in it's area class either. And perf/area is by far Skymont's best attribute.

Yeah of course, but at least that will be a reset. If Coyote Cove sucks, so will Griffin Cove in RZL, so the unified core would be the next earliest chance at not sucking in ST.

I do think at the very least the E-core team has done better than the P-core team in recent years, so I have more hope for them. Arctic Wolf will be one to keep an eye on IMO, those Atom team cores have carried the -mont moniker for 15 years so the shift towards a new codename structure indicates that will be the first real basis for the future Unified Core.

 

[OneEng2]

While others seem to have addressed this question rather well, it seems that (for now) Intel hasn't progressed much IPC-wise with thier P cores over the last few generations. Combine that with the rumours that Nova Lake/Coyote Cove won't see major (or any) P-core clockspeed updates, and you'll have a situation where Nova Lake will struggle with ST performance (as well as MT performance, particularly in the range of 8-16 threads).

Well if you weren't allergic to a "wall of text" you could explain to us, in detail, where and why I'm wrong. But nope, with you, its "trust me bro". That shtick has gotten old.

PTL has fixed the L3

How is the latency now?

It's the least performant N3 Node out of the family

Link please.

 

[hemedans]

Yea... no. Windows laptops routinely achieve Macbook like idle power. It's when they actually need to compute (SoC, CPU), AMD and Intel fall apart.

Lunar lake is exactly like that good at low power and lose at high power, under 15W is better than competition and you can even game with it under 10W, M5 Macbook would use significant more power than LNL when playing AAA games.

 

[poke01]

Lunar lake is exactly like that good at low power and lose at high power, under 15W is better than competition and you can even game with it under 10W, M5 Macbook would use significant more power than LNL when playing AAA games.

You do know that M5 MacBook also has a low power mode and operates at <10watts then as well.
Look lunar lake is great and all but it’s no where comparable to M5.

The point is M5 can do both (it can be very low power and powerful) meanwhile Lunar Lake is just an efficient SoC, there’s a reason why Intel made Panther Lake to combine best of Lunar lake and arrow lake.