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.

 

[adroc_thurston]

Peak clocks won’t regress much, if at all over raptor lake.

They really did!
And it's only just the beginning.

Finally (to no one in particular), what is up with you folks throwing peak per core power numbers around?

It kinda does matter in predominantly 1t web browsing loads.

Also, race to idle is a thing.

Joules spent per workload total also matter.

 

[TESKATLIPOKA]

CPU performance uplift will mostly be multicore. Perf/watt much improved.

Even If singlecore sees only 5% improvement, It's not a big deal for me. It's not like AMD has an edge here, at least until Strix Point.
MT is more important here. Raptor was not doing that great in multithread at limited TDP despite having more cores, because of low clocks.
Now It looks like we will see a significant increase in clocks resulting in much better MT at limited TDP.
Then there is still the IGP.

 

[Schmide]

Finally (to no one in particular), what is up with you folks throwing peak per core power numbers around? That number is kind of meaningless. A single core on a Zen 4 chip may consume 20W, but you will never see it consume 20W in an all-core workload.

A well-designed CPU should be able to scale as much as possible as long as power limits haven’t been reached. If the power limit is 28W there is absolutely nothing wrong with a single core using nearly all of it.

Also, race to idle is a thing. As long as chip a can complete tasks quicker and use less total power than chip b, chip a is the winner, even if chip a peaks at 28W per core vs chip B’s 5-10W per core.

Yeah. Wait ? Are we not supposed to talk about peak per core power like you just did ?

Nothing happens in a vacuum. Different algorithms and data cause different areas of the processor to do work. A large FFT may fill the cache and spill into main memory over and over, while an iterative matrix model may comfortably reside in the L1.

Cores exist in a hierarchy and workloads vary. You can win all categories, some, or none. Even if chip a wins against chip b, there is probably a low power variant of chip a or b that wins the power metric regardless of time. Conversely there is probably a server class chip that takes power to performance to a whole new level, and it sucks at idling.

 

[eek2121]

They really did!
And it's only just the beginning.

We will see. The source I got the boost clocks from is a reputable one.

Remember these are i5/i7 chips, not i9s.

The Core i7-1370p had a boost clock of 5.2 ghz. The 1360p boosts to just 5.0 ghz.

Intel is positioning Meteor Lake as a premium chip (hence the “ultra” moniker), they would not be doing that unless these parts are beating non “ultra” parts.

 

[Khato]

Yeah, MTL performance will definitely be on par or greater than RPL while being more efficient.

It's plausible that notable clock speed regression is with respect to the e-cores rather than p-cores.

 

[msj10]

Peak clocks won’t regress much, if at all over raptor lake. The last I saw these chips would have a peak of around 4.8-4.9ghz, but that has probably changed.

tbh this is still a big regression from the 5.4Ghz in RPL-H

 

[mikk]

If it can't hit 5.0 Ghz 1T it's a little disappointing.

 

[adroc_thurston]

If it can't hit 5.0 Ghz 1T it's a little disappointing.

It is what it is.
Still a nice change for Intel.

 

[Exist50]

If it can't hit 5.0 Ghz 1T it's a little disappointing.

It'll hit 5GHz. But certainly MTL won't be noteworthy from a CPU perspective.

 

[H433x0n]

If it can't hit 5.0 Ghz 1T it's a little disappointing.

There’s a decent probability that single core boost clock will get to 5.0-5.2ghz.

OneRaichu has had a MTL-H ES2 sample up to 5.1ghz.

 

[adroc_thurston]

OneRaichu has had a MTL-H ES2 sample up to 5.1ghz

He has nothing and the less you listen to him, the merrier.

 

[Exist50]

He has nothing and the less you listen to him, the merrier.

On the contrary, Raichu has been a pretty reliable leaker.

 

[adroc_thurston]

On the contrary, Raichu has been a pretty reliable leaker.

He just repeats whatever he sees on Weibo, a classical Chinese tourist in every way possible.

 

[msj10]

He just repeats whatever he sees on Weibo, a classical Chinese tourist in every way possible.

this is total BS and you know it
did he also get this slide from Weibo? lol

 

[H433x0n]

He just repeats whatever he sees on Weibo, a classical Chinese tourist in every way possible.

That’s more than I can do since I can’t read mandarin nor access that market. The people I know at Intel have a limited scope of knowledge that’s based around their current project.

 

[adroc_thurston]

did he also got this slide from Weibo? lol

Yea a ton of stuff gets shared there.

The people I know at Intel have a limited scope of knowledge that’s based around their current project. In that sense Raichu helps fill the gaps for me with the client market.

Yea fine but quoting him as a 'leaker' is silly.

 

[Exist50]

He just repeats whatever he sees on Weibo, a classical Chinese tourist in every way possible.

Then by all means post the original source(s) if he's just parroting things. Not that it matters practically either way.

 

[Khato]

this is total BS and you know it
did he also get this slide from Weibo? lol

Did the full slide get leaked, or just the redacted version with MTL-S removal? Just one little tidbit from an interesting slide.

 

[TESKATLIPOKA]

It'll hit 5GHz. But certainly MTL won't be noteworthy from a CPU perspective.

We will see when It's released. From 1T I expect the same performance, MT being better. The question is how much better.

4P+8E i7-1360P 30W - 11571 points (100%)
8C16T R7 PRO 6860Z 27W - 11822 points (102%)
6P+8E i7-1370P 30W - 13157 points (114%)
8C16T Ryzen Z1 Extreme 30W - 14960 points (129%)

This is how i7-1360P looks during CB R23 MT.

At 35-40W P-core clocks at ~3.05GHz and E-core at 2.3-2.4GHz.

If 4P+8E MTL at 28-30W can't score at least the same as Phoenix 8C16T, which I consider an equal, then It's really not interesting.

 

[mikk]

It'll hit 5GHz. But certainly MTL won't be noteworthy from a CPU perspective.

This was my expectation but what is the big deal about then? We know since ages that MTL will bring a sidegrade at best when it comes to 1T performance, nobody expects RPL boost clocks. I would expect at least ADL boost clocks though, means at least 5.0 Ghz (i9-12900H). adroc_thurston implies it's really bad like catastrophic. If it hits at least 5 Ghz it's not bad to me.

I expect much better PL1 clocks however because of Intel 4, efficiency is the key there. Otherwise it would be a disappointment...

He just repeats whatever he sees on Weibo, a classical Chinese tourist in every way possible.

I don't think so, sorry. You could prove me wrong but I doubt you can.

 

[eek2121]

tbh this is still a big regression from the 5.4Ghz in RPL-H

This is a P chip. Intel will also rebadge these as H chips.

There is at least one ES chip that reportedly hit 5.4ghz.

This was my expectation but what is the big deal about then? We know since ages that MTL will bring a sidegrade at best when it comes to 1T performance, nobody expects RPL boost clocks. I would expect at least ADL boost clocks though, means at least 5.0 Ghz (i9-12900H). adroc_thurston implies it's really bad like catastrophic. If it hits at least 5 Ghz it's not bad to me.

I expect much better PL1 clocks however because of Intel 4, efficiency is the key there. Otherwise it would be a disappointment...




I don't think so, sorry. You could prove me wrong but I doubt you can.

The reason I suspect this chip will be a bit of a game changer for Intel is that it is low power. PL1 clocks went up massively to 3.4ghz despite a 28W PL1, and if it does indeed have 2 cores on the SoC, most of the chip can power down during idle/low loads. I could easily see battery life doubling over previous-gen if Intel nails execution.

AMD needs to take a page out of Intel’s book. If they put a couple cores on the Ryzen IOD it would do wonders for idle power consumption.

 

[mderbarimdiger]

​

Yea a ton of stuff gets shared there.

you have made a ton of grand claims in the last week period, maybe they are right, but i cant verify them, your claims about A20 non existent will be proved or disproved in due time, to early to judge, your claims about having knowledge about Turin is something i cant prove, you made some other bold statements but its hard to prove them or disprove them, however this one accusation you have made is something you should be able to prove right here and right now.

please prove it. link to the weebo slide please?

 

[A///]

On the contrary, Raichu has been a pretty reliable leaker.

i lost track of what post of yours i wanted to quote but i'd point mtl being like tigerlake on the scale but better.

 

[uzzi38]

There’s a decent probability that single core boost clock will get to 5.0-5.2ghz.

OneRaichu has had a MTL-H ES2 sample up to 5.1ghz.

Utter bull.

MTL-H ES2 doesn't even get close.

 

[uzzi38]

​

you have made a ton of grand claims in the last week period, maybe they are right, but i cant verify them, your claims about A20 non existent will be proved or disproved in due time, to early to judge, your claims about having knowledge about Turin is something i cant prove, you made some other bold statements but its hard to prove them or disprove them, however this one accusation you have made is something you should be able to prove right here and right now.

please prove it. link to the weebo slide please?

Raichu being a Weibo resident and often just reposts stuff to Twitter has been known for ages now, this isn't some new revelation. I know for sure that I've said as much here before as well.

He does occasionally post slides and stuff that are legit, but there are also times where there are huge misses. What he's said about MTL is one of those cases.