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.

 

[MoistOintment]

Interesting, I do hope that Intel fixes that issue of the brightness getting lowered on Lunar Lake especially on the IPS models. It’s efficient enough that there is no need to that anymore.

On MacBooks the display only dims on low power mode setting is enabled.

My experience with Windows 11 brightness behavior is that it remembers your brightness setting when plugged and unplugged, and sets the brightness back to where you last left it in that state when you switch between states.

If you have max brightness when plugged, and you unplug, the brightness level will switch to the brightness level you were using when you were last unplugged, whatever that may be.

 

[adroc_thurston]

Halo is up to 16 performance cores with chiplets

That's not the selling point.

Definitely sacrifices battery life to get even more performance over standard Strix.

That's Fire Range.

It’s also meant to be used an APU for handheld devices.

That's the reason why it's a one-off to begin with. Tablet parts are extremely niche.

 

[jpiniero]

That's not the selling point.

That would be AI.

 

[adroc_thurston]

That would be AI.

No lmao.

 

[Josh128]

MT is not that important compared to ST in thin and lights. Apple has already shown that to the world with its Apple Silicon. Lunar Lake is just following the same path. AMD not so much.

AMD Strix is will remain a disappointment this gen.


Agree.

Whoa, where are all these supposed Lunar Lake ST perf wins over Strix?! Outside of a notable CB R24 win, they are few and far between. It loses or ties in R23 & Geekbench, loses in R20 and R15. Have you seen the battery life in the Strix Vivobook S 14? Its right up there with Lunar Lake, lol. A lot of people are ignorant of the fact that the screen size and type affect battery much more than these new gen CPUs. Almost every single Lunar Lake review Ive seen has compared battery life against a Strix with a 2" larger OLED screen, which make all the difference in the world in battery life. Its a completely apples to oranges comparison. What exactly is disappointing about Strix to you?

Review: Intel Lunar Lake CPUs combine good battery life and x86 compatibility

But it’s too bad that Intel had to turn to TSMC to make its chips competitive.

arstechnica.com

Vivobook S 14

 

[LightningZ71]

Wrong.

N7 is a crappy node compared to Intel 7. Intel 7 is far ahead. Plus, it offers better performance than N7.

TSMC N7 and Intel 7 (10esf+) are broadly equivalent in many ways, but, when it came down to implementation, N7 had higher PRACTICAL density. For Intel to achieve the performance that they needed from Alder Lake to Raptor lake refresh, they continually relaxed their logic density. N7 products of equivalent performance had better density in production form. N6, being an N7 fanily node with very little logic density improvement was so much better than Intel’s volume node, Intel 7, that Intel chose to outsource several products to it over using their own in-house node.

 

[branch_suggestion]

TSMC N7 and Intel 7 (10esf+) are broadly equivalent in many ways, but, when it came down to implementation, N7 had higher PRACTICAL density. For Intel to achieve the performance that they needed from Alder Lake to Raptor lake refresh, they continually relaxed their logic density. N7 products of equivalent performance had better density in production form. N6, being an N7 fanily node with very little logic density improvement was so much better than Intel’s volume node, Intel 7, that Intel chose to outsource several products to it over using their own in-house node.

Indeed, Intel clocks higher at the cost of density. And recent TSMC nodes can now achieve similar clocks with less density penalty.
Not to mention 18A was downgraded from what was previously expected, so it is now N3P PPA overall instead of N2.
Not to mention the cost and throughput are worse than equiv TSMC nodes, yields are also worse but serviceable.

 

[jdubs03]

Indeed, Intel clocks higher at the cost of density. And recent TSMC nodes can now achieve similar clocks with less density penalty.
Not to mention 18A was downgraded from what was previously expected, so it is now N3P PPA overall instead of N2.
Not to mention the cost and throughput are worse than equiv TSMC nodes, yields are also worse but serviceable.

I was listening to Ian’s podcast the other day (it was TechPoutine 6) about 18A and the performance improvements vs. 20A and Intel 3, courtesy of a question @cebri1 asked, and he basically said that he considers what Intel said early on in the year (their foundry day) about their process characteristics. So seemingly wait for additional info about 18A and its performance improvement vis-a vis Intel 3.

That’s not to say that the reduction expectations is not going to be reality. I think there’s a better chance that it will be. But, there’s still a chance that it won’t be.

 

[511]

LNL has same MT throuhgput and efficency as a 6C/12T Phoenix 2 7540U, i would expect in a 1400€ laptop better MT perf and perf/Watt than laptops sold at 600-650€,
more than 2x the price at same MT perf just because of the GPU.?.

It's single core is way higher like 20-30%
Battery life is better
Gpu is way better

That s the magic of the marketing, actual numbers are not even considered, to the point that a CPU that should be in the sub 1000€ laptop market is accepted as being valuable at 1400€ if not 2000€.

MacBook should not exist by this logic than 🤣

 

[Abwx]

It's single core is way higher like 20-30%

Lol, with beliefs like this i understand better the hype.

Compare LNL to the 7840U that has about same ST perf than a 7540U/7640U.
and tell us where are those 20-30%, 4.8% in GB 6 FTR.

Review: Intel Lunar Lake CPUs combine good battery life and x86 compatibility

But it’s too bad that Intel had to turn to TSMC to make its chips competitive.

arstechnica.com

 

[poke01]

Whoa, where are all these supposed Lunar Lake ST perf wins over Strix?! Outside of a notable CB R24 win, they are few and far between. It loses or ties in R23 & Geekbench, loses in R20 and R15. Have you seen the battery life in the Strix Vivobook S 14? Its right up there with Lunar Lake, lol. A lot of people are ignorant of the fact that the screen size and type affect battery much more than these new gen CPUs. Almost every single Lunar Lake review Ive seen has compared battery life against a Strix with a 2" larger OLED screen, which make all the difference in the world in battery life. Its a completely apples to oranges comparison. What exactly is disappointing about Strix to you?

Review: Intel Lunar Lake CPUs combine good battery life and x86 compatibility

But it’s too bad that Intel had to turn to TSMC to make its chips competitive.

arstechnica.com

Vivobook S 14

Wow that Ars review is an eye opener

Intel designed Lunar lake to be only good in thin and light use cases and it shows in this benchmark. Its not made for MT at all, Intel will offer ARL H instead for that.




Intel targeted the ultrabook crowd who want the best x86 battery life and there is a market for that. In video playback and web browsing, Lunar lake will win all battery tests due to its PMIC, Skymont core and overall arch that is effeciency focused. Doing high performance tasks that put the whole cores under load is not what Lunar Lake was designed for.

 

[511]

That would be AI.

Please No AMD would loose hard in client 🤣 they are living of Intel libraries like OpenVino to get performance 🤣🤣

 

[cannedlake240]

I was listening to Ian’s podcast the other day (it was TechPoutine 6) about 18A and the performance improvements vs. 20A and Intel 3, courtesy of a question @cebri1 asked, and he basically said that he considers what Intel said early on in the year (their foundry day) about their process characteristics. So seemingly wait for additional info about 18A and its performance improvement vis-a vis Intel 3.

That’s not to say that the reduction expectations is not going to be reality. I think there’s a better chance that it will be. But, there’s still a chance that it won’t be.

They would've clarified it already by now if it was just a mistake. They bet the company on that node, no way they're allowing typos to influence investor confidence.

 

[511]

Indeed, Intel clocks higher at the cost of density. And recent TSMC nodes can now achieve similar clocks with less density penalty.

No? Why is Intel 4/3 so close to N3 with their 3-3 HPC Library in PPA
2-2 is relatively less dense but better performant than their 2-2 is and also Nanoflex to widen it due to 2-1 fin

Not to mention 18A was downgraded from what was previously expected, so it is now N3P PPA overall instead of N2.

I am still expecting between N3P and N2.
N3P on a worse scenario N2 in a good one but it is available on same time as N3P N2 is a year late 🙂

Not to mention the cost and throughput are worse than equiv TSMC nodes, yields are also worse but serviceable.

They have corrected the cost with Intel 3 yield are on Par with "10nm+++" Intel 7 you can read Tech insights report

 

[511]

Wow that Ars review is an eye opener

Intel designed Lunar lake to be only good in thin and light use cases and it shows in this benchmark. Its not made for MT at all, Intel will offer ARL H instead for that.
View attachment 108298

View attachment 108299

Intel targeted the ultrabook crowd who want the best x86 battery life and there is a market for that. In video playback and web browsing, Lunar lake will win all battery tests due to its PMIC, Skymont core and overall arch that is effeciency focused. Doing high performance tasks that put the whole cores under load is not what Lunar Lake was designed for.

You know we have quickSync as well to destroy this task with more and wider codec than any of the mentioned

 

[jdubs03]

They would've clarified it already by now if it was just a mistake. They bet the company on that node, no way they're allowing typos to influence investor confidence.

Yeah, I mean I did say that it doesn’t seem like a mistake. But he’s in the know so it is a valid opinion. It is a shame that it’ll take 4-5 more months til their next foundry day.

 

[poke01]

You know we have quickSync as well to destroy this task with more and wider codec than any of the mentioned

True I was just assessing CPU performance in an all core load

 

[H433x0n]

Whoa, where are all these supposed Lunar Lake ST perf wins over Strix?! Outside of a notable CB R24 win, they are few and far between. It loses or ties in R23 & Geekbench, loses in R20 and R15. Have you seen the battery life in the Strix Vivobook S 14? It’s right up there with Lunar Lake, lol. A lot of people are ignorant of the fact that the screen size and type affect battery much more than these new gen CPUs. Almost every single Lunar Lake review Ive seen has compared battery life against a Strix with a 2" larger OLED screen, which make all the difference in the world in battery life.

Lion Cove and Zen 5 perform within 1-2% of each other, with Lion Cove getting a 1-2% edge in integer performance as well as efficiency going off of David Huang’s data. The 1T performance is going to be completely dependent on the laptop OEM and the chassis it is in. For all intents and purposes, the 1T performance is the same.

Its a completely apples to oranges comparison. What exactly is disappointing about Strix to you?

Agreed, it is apples to oranges. That’s why trying to compare nT performance makes no sense. Despite the defensiveness of multiple posters, nobody is saying Strix is disappointing.

 

[511]

Sigh RWC PPA

 

[cebri1]

Sigh RWC PPA View attachment 108301

You conveniently ignore the next paragraph which says it also performs 12% faster putting it at parity ppw.

Again, RWC is going to lose against Turin, but Intel has gone from having a processor that was 40% slower than Genoa X while consuming more power to achieve ppw parity in 24 months. Turin will put them in the lead again but before it was a bloodbath.

Edit: btw this is against Genoa-X, I don’t think we’ll see vcache server parts until next year. Against Genoa, GNR has a 23% performance lead (not ppw).

 

[cannedlake240]

You conveniently ignore the next paragraph which says it also performs 12% faster putting it at parity ppw.

Again, RWC is going to lose against Turin, but Intel has gone from having a processor that was 40% slower than Genoa X while consuming more power to achieve ppw parity in 24 months. Turin will put them in the lead again but before it was a bloodbath.

Both process and core just aren't quite there yet to match TSMC + zen. GNR 3.2ghz all core is quite low

 

[AcrosTinus]

You conveniently ignore the next paragraph which says it also performs 12% faster putting it at parity ppw.

Again, RWC is going to lose against Turin, but Intel has gone from having a processor that was 40% slower than Genoa X while consuming more power to achieve ppw parity in 24 months. Turin will put them in the lead again but before it was a bloodbath.

You are right, though I won't make any claims about Turin. Intel had the memory subsystem advantage and on the next Epyc it won't change. Furthermore Zen5 is not a generation faster than 8Wide RWC...
And even if Turin turns out to be faster "no pun intended", they don't have the software stack and accelerators that Intel has.

 

[cebri1]

You are right, though I won't make any claims about Turin. Intel had the memory subsystem advantage and on the next Epyc it won't change. Furthermore Zen5 is not a generation faster than 8Wide RWC...
And even if Turin turns out to be faster "no pun intended", they don't have the software stack and accelerators that Intel has.

I think Turin will have a considerable lead in efficiency. Performance wise I’m not so sure. There is a significant gap to cover.

 

[jdubs03]

You conveniently ignore the next paragraph which says it also performs 12% faster putting it at parity ppw.

Again, RWC is going to lose against Turin, but Intel has gone from having a processor that was 40% slower than Genoa X while consuming more power to achieve ppw parity in 24 months. Turin will put them in the lead again but before it was a bloodbath.

I think Turin will have a considerable lead in efficiency. Performance wise I’m not so sure. There is a significant gap to cover.

This is why Clearwater Forest and Diamond Rapids are so important. If they can release these on time next year on 18A, plus Panther Lake, they’ll be able to turn the tide in the x86 space and moat themselves in Windows against Qualcomm.
I admit I’m a firm believer that there is light at the end of the tunnel. Now for IPC leadership, that’s unrealistic.

 

[cebri1]

This is why Clearwater Forest and Diamond Rapids are so important. If they can release these on time next year on 18A, plus Panther Lake, they’ll be able to turn the tide in the x86 space and moat themselves in Windows against Qualcomm.
I admit I’m a firm believer that there is light at the end of the tunnel. Now for IPC leadership, that’s unrealistic.

diamond rapids is 2026. It will compete against Zen6. CWF I agree it could shake the current landscape but we’ll see.