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.

 

[Win2012R2]

It's literally their best yielding product in the last 10 years

I had 5 Ghz all cores 9900K on 14nm, and this is what - 5-5.1 Ghz (few cores boost only?) on 1.8nm

 

[jpiniero]

I had 5 Ghz all cores 9900K on 14nm, and this is what - 5-5.1 Ghz (few cores boost only?) on 1.8nm

Probally going to be meme level volume. So they are being conservative on clocks.

 

[fastandfurious6]

5ghz was easily reachable since sandy bridge even no?

32nm?

sandy bridge was notorious for easy unlock +20% perf etc, was a superchip

edit: yeah quick search shows posts from 2020 saying "bro my sandy 2600k @5ghz still runs everything I throw at it just fine, no reason to upgrade"

 

[511]

I had 5 Ghz all cores 9900K on 14nm, and this is what - 5-5.1 Ghz (few cores boost only?) on 1.8nm

14nm++++ vs 18A there is not even a single + in the node

Probally going to be meme level volume. So they are being conservative on clocks.

5.1 GHz SKU is going to be like 185H even with LNL the most seen SKU is 256/258V.

 

[Win2012R2]

So they are being conservative on clocks.

Intel?

14nm++++ vs 18A there is not even a single + in the node

14nm vs 1.8nm, that's:

14nm -> 10nm -> 7nm -> 4nm -> 3nm -> 1.8nm

One would have thought clocking 4 (p)cores will be easier than 6-8 too.

 

[511]

14nm -> 10nm -> 7nm -> 4nm -> 3nm -> 1.8nm

10nm/7nm and 4/3 are the same node family for Intel.

 

[igor_kavinski]

It can possibly clock higher but the heat output must be getting out of hand. No point in shipping a CPU that immediately throttles with the kind of cooling they have in typical laptops.

 

[511]

It can possibly clock higher but the heat output must be getting out of hand. No point in shipping a CPU that immediately throttles with the kind of cooling they have in typical laptops.

Only a review will tell us.

 

[Win2012R2]

It's literally their best yielding product in the last 10 years

How can you possibly know that?

 

[igor_kavinski]

Only a review will tell us.

Laptop chips can't usually be overclocked so we need to wait until at least an HX laptop review and if that even has an option to overclock using XTU or its own BIOS.

But those industrial mobos with PTL may have a clue...

 

[Josh128]

How is it Cannon Lake/Ice Lake 2.0? Those didn't even go over 4 GHz even on a 4 wide core this is 5 GHz 8 wide LNC+ it's EEP Intel has done this for like quite some time.

Low clocks, low yields, limited launch primarily to satisfy promises to investors.

 

[511]

How can you possibly know that?

Same way I have the WCL PPW Chart

 

[511]

Low clocks, low yields, limited launch primarily to satisfy promises to investors.

The yields are not low tbh and it's always limited same for LNL real volume is Q1 26.
Personally I was expecting clocks to be 5.2-5.3 Ghz

 

[Josh128]

The yields are not low tbh and it's always limited same for LNL real volume is Q1 26.

Well, it is literally the only product coming out of the much ballyhooed 18A process, and its a smallish tile, so I guess thats feasible, due in no small part to it being a chiplet. Ice Lake was 130mm2 monolithic, so ya.

 

[511]

Well, it is literally the only product coming out of the much ballyhooed 18A process, and its a smallish tile, so I guess thats feasible, due in no small part to it being a chiplet. Ice Lake was 130mm2 monolithic, so ya.

OFc it is There were only 2 products scheduled to launched for 2025 on 18A one of them got delayed so PTL is all that left feels kind of bad for Clearwater forest.

 

[Win2012R2]

Same way I have the WCL PPW Chart

What's the current defect rate for 18A and what is your source for it?

 

[511]

What's the current defect rate for 18A and what is your source for it?

Bruh really Source ? And D0 you meant the physical D0 right It's less than <0.2

 

[igor_kavinski]

What's the current defect rate for 18A and what is your source for it?

He can't provide any source or he will get fired

 

[511]

He can't provide any source or he will get fired

That is public info

 

[Philste]

I don't know why everyone is going crazy with the ~5GHz clocks. There looks to be a ceiling around that Point for mobile CPUs, some reused Desktop Dies in AMD/Intel HX series do a bit more, but otherwise...

AMD had 4.8GHz in 2021 with Cezanne, now 4 Years later Strix (Halo) does 5.1GHz.

With Intel its kinda the same Story, just with a few MHz higher. Their performance SKUs have been around 5.4 GHz for years now, the more efficient ones are at these 5-5.1 GHz, no matter if N3B, Intel 4 or 18A.

Qualcomm is also slowly running against a wall now. First 8 Elite Gen 5 tests show 12W power draw for Single Core loads, in a phone SoC, thats more than the max overall power draw we had for these products like 2-3 years ago. Pretty sure that 5GHz on N3X cores (X Elite Gen2 Extreme) will use 25W+, thats no better than x86 cores.

I doubt any Medusa Mobile product will go beyond 5.4/5.5GHz, despite Desktop ZEN6 SKUs going 1GHz higher. There just seems to be a Point around 5 GHz, that doesn't make Laptop SKUs over it sufficient.

 

[gdansk]

despite Desktop ZEN6 SKUs going 1GHz higher.

You should press X to Doubt that too

 

[511]

You should press X to Doubt that too

Going 6 GHz is pretty easy with N2

 

[511]

Qualcomm is also slowly running against a wall now. First 8 Elite Gen 5 tests show 12W power draw for Single Core loads, in a phone SoC, thats more than the max overall power draw we had for these products like 2-3 years ago. Pretty sure that 5GHz on N3X cores (X Elite Gen2 Extreme) will use 25W+, thats no better than x86 cores

It's 18-20W if you take Qcoms word for it

With Intel its kinda the same Story, just with a few MHz higher. Their performance SKUs have been around 5.4 GHz for years now, the more efficient ones are at these 5-5.1 GHz, no matter if N3B, Intel 4 or 18A.

Intel's '+' node goes to 5.3-5.4 GHz on mobile without blowing up.

 

[Magio]

I don't know why everyone is going crazy with the ~5GHz clocks. There looks to be a ceiling around that Point for mobile CPUs, some reused Desktop Dies in AMD/Intel HX series do a bit more, but otherwise...

I don't think we should want more than 5GHz in a vacuum, but the stagnant IPC with a growing IPC deficit to the ARM space means that without higher clocks there's little to no performance uplift. Like a new node with BSPD + GAAFET proudly delivering worse than Arrow Lake ST performance is just bad.

 

[511]

I don't think we should want more than 5GHz in a vacuum, but the stagnant IPC with a growing IPC deficit to the ARM space means that without higher clocks there's little to no performance uplift. Like a new node with BSPD + GAAFET proudly delivering worse than Arrow Lake ST performance is just bad.

There is only PPW uplift from the node Intel never gave up Tick Tock model they still use it for derisking.