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

Page 131 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Tigerick

Senior member
Apr 1, 2022
831
794
106
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 - NIntel Wildcat LakeIntel Lunar LakeMediatek D9500
Launch DateQ1-2023Q2-2026 ?Q3-2024Q3-2025
ModelIntel N300?Core Ultra 7 268VDimensity 9500 5G
Dies2221
NodeIntel 7 + ?Intel 18-A + TSMC N6TSMC N3B + N6TSMC N3P
CPU8 E-cores2 P-core + 4 LP E-cores4 P-core + 4 LP E-coresC1 1+3+4
Threads8688
Max Clock3.8 GHz?5 GHz
L3 Cache6 MB?12 MB
TDP7 WFanless ?17 WFanless
Memory64-bit LPDDR5-480064-bit LPDDR5-6800 ?128-bit LPDDR5X-853364-bit LPDDR5X-10667
Size16 GB?32 GB24 GB ?
Bandwidth~ 55 GB/s136 GB/s85.6 GB/s
GPUUHD GraphicsArc 140VG1 Ultra
EU / Xe32 EU2 Xe8 Xe12
Max Clock1.25 GHz2 GHz
NPUNA18 TOPS48 TOPS100 TOPS ?






PPT1.jpg
PPT2.jpg
PPT3.jpg



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.



LNL-MX.png
 

Attachments

  • PantherLake.png
    PantherLake.png
    283.5 KB · Views: 24,028
  • LNL.png
    LNL.png
    881.8 KB · Views: 25,522
  • INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg
    INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg
    181.4 KB · Views: 72,430
  • Clockspeed.png
    Clockspeed.png
    611.8 KB · Views: 72,318
Last edited:

dullard

Elite Member
May 21, 2001
25,989
4,598
126
There have been a series of Intel comments that are close to what Abwx said over the last couple of years. Just search for the term "unquestioned leadership" which Intel has been using repeatedly as their goal over the next few years.
The rest is Abwx's own interpretation of those statements.

Do note that there is a significant difference between "transistor performance per watt" and "CPU performance per watt". You could have a great process making great transistors but still have a horrible CPU design. Having a great transistor helps, but is not sufficient. People here like to blur the lines to make personal statements.
 
Last edited:

SiliconFly

Golden Member
Mar 10, 2023
1,924
1,284
106
But i did read the article and looked at Intel s slides for expected perf/watt improvement, i didnt pull the 52% perf/watt improvement out of nowhere, i let hasardous estimations for whom think that they know better than Intel themselves.

Beside they state that ll regain perf/watt leadership in 2025, so they are admitting that they dont hold it currently, yet there s people who apparently also know better than Intel in this area...
Pat said AMD will be in the rear view mirror in 2025. I don't fully agree with him, cos it's way too optimistic and more of a marketing pitch. But what he actually tried to convey was that Intel will lead in all aspects in client starting 2025 (his words and his words alone).

But that doesn't mean Intel can't have an ppw advantage next year with ARL. Intel may or may not gain ppw advantage next year with ARL. ARL might easily match Zen 5 in ppw when they both launch next year. It's very much possible. There is a slight chance ARL might even edge out Zen 5 a bit in ppw next year. That too actually is a possibility.
 

Abwx

Lifer
Apr 2, 2011
11,879
4,864
136
There have been a series of Intel comments that are close to what Abwx said over the last couple of years. Just search for the term "unquestioned leadership" which Intel has been using repeatedly as their goal over the next few years.
The rest is Abwx's own interpretation of those statements.

Do note that there is a significant difference between "transistor performance per watt" and "CPU performance per watt". You could have a great process making great transistors but still have a horrible CPU design. Having a great transistor helps, but is not sufficient. People here like to blur the lines to make personal statements.




Intel is today defining 'technology leadership by 2025' as defined by the metric of performance per watt.

My own interpretation did you say..?.

 

dullard

Elite Member
May 21, 2001
25,989
4,598
126
Nowhere there does it say they are behind today though. They aren't even talking about desktop anyways. Of course they are behind in the server space.
But in Abwx's world, regaining a lead means admitting you are losing. It is like a hockey/baseball/soccer game. Suppose team X was up 1-0, then team Y scored twice to be up 1-2, and now team X scored to make it a tie at 2-2. If team X gets another goal to regain the lead, it means that team X "admits" they are losing at a score of 2-2. It is an interesting interpretation of the data, and he'll back it up by posting unsourced quotes and links that are not directly related to the discussion.
 
Last edited:

jpiniero

Lifer
Oct 1, 2010
16,780
7,230
136
I seriously doubt it. In spite of all the rumors suggesting that Intel might use TSMC N3B for the CPU tile, I believe Intel is gonna stick to Intel foundry nodes for their future CPU tiles.

The options for Intel might be either go ahead with using N3B or sell Raptor Lake Refresh Refresh on desktop. IMO they can't hide the volume on desktop like they can on mobile.

Edit: I suppose they could port the Arrow Lake tile to some other node, even possibly 10 nm.
 

Geddagod

Golden Member
Dec 28, 2021
1,521
1,615
106
And that's a problem they need to fix, how is that relevant? AMD doesn't have to deal with 10 nm but instead use TSMC's 5. That's an intel problem that they need to fix. Why would you even bring that up?
Because
a)AMD doesn't deal with the IO die in mobile, where they have monolithic chips, and they poop all over Intel's competing mobile chips in that segment as well
b)AMD is fine with their IO die setup on desktop, because no one cares about how well the 7950x or 13900k performs at 45 watts, outside of reviewer's interest as a theoretical.
Neither of those are true though. There are plenty (actually, the vast majority) of applications and games that don't draw over 40-50-60 watts. Even creativity applications like the whole adobe suite etc do not use that much power. 99% of day to day tasks Intel is both faster and more efficient.
But those don't boost all cores at 45-50-60 watts. They boost a couple cores high. Those applications are much less well threaded, and are subject to different boosting algorithms, and also way less likely to rely on cross CCD communication as well.

Also, how can idle power be a draw? You just said AMD has the IO die that consumes power, that's why it doesn't scale properly at 45w. So how does it idle at the same levels as Intel? That doesn't make sense to me.
As I reiterated many times, looking at all core power draw at 45 watts is much different than looking at idle. AMD has probably optimized power draw and how much each section of the CPU (soc/core) would get latter, not so much for the former. Idle power draw is pretty much identical for AMD and Intel across the system:
1697500325985.png
Intel has the most efficient desktop CPU's.
At a level of power consumption no one cares about or matters. AMD has the most power efficient chips for the vast majority of use cases.
But again, regardless of that, the point is most people wouldn't be running 24/7 rendering, in most tasks that people use the Intel is more efficient due to the uncore burden you just mentioned.
If you're buying a 13900k or 7950x, you almost certainly will be using all those cores, yes. And most of DIY would also be going for gaming as well, which both the 3D and vanilla 7950x are more efficient than the 13900k in
And rumors about 3d cache. Imagine the 13900k with 3d cache....
There are none. At best there's the Adamantine cache rumors (which don't exist for ARL lol), but seeing how ARL doesn't use Foveros Direct, the latency between the two chiplets is going to be much higher, and not at all comparable to the 3D stacking of V-cache.
But all is well, The current desktop 14th gen is RPL refresh based on Intel 7. Next desktop 15th gen is Arrow Lake based on Intel 20A with a massive PPW gain of 63% when compared to 14th gen RPL refresh
I don't know how many times I have to say this, but you can not simply multiply all the perf/watt values together. I used to think that as well, I'll admit, but it's nonsensical since we don't know where Intel is measuring the perf/watt gains at each point of the curve.
You have too big expectations, according to Intel themselves the used process will be at best 52% more efficient at same throughput and same uarch...
Where tf are you getting those numbers from?
i didnt pull the 52% perf/watt improvement out of nowhere,
No, I feel like you are lol ....
I suppose they could port the Arrow Lake tile to some other node, even possibly 10 nm.
14mm^2 per core lol
so ARL could eventually be comparable as current AMD products on the DT side of thing but not for mobile,
Lmao the cope here is wild
 

Hulk

Diamond Member
Oct 9, 1999
5,138
3,726
136
Both AMD/Intel top-of-the-stack parts are pretty close in terms of overall performance. But when it comes to efficiency there is no comparison. In most cases Zen 4 is over TWICE as efficient!

I still love my 13600K but those Raptor Coves do like electricity!

1697503906357.png
 

cebri1

Senior member
Jun 13, 2019
373
405
136
unless you are trolling are you comparing an 8+16 processor with a 6+8?

The numbers seem wrong because the MT score is significantly below the 155H. The ST score is matching the 13700HX with 2 p-cores less and 6% less max frequency.
 
  • Like
Reactions: Henry swagger
Jul 27, 2020
27,709
18,972
146
unless you are trolling are you comparing an 8+16 processor with a 6+8?
13900H is also 6+8.


Intel will probably sell the Ultra for higher than the 13900H's price. But yeah if the battery life is great, then I can see the higher price justified.
 
  • Like
Reactions: Tlh97 and cebri1

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
There are none. At best there's the Adamantine cache rumors (which don't exist for ARL lol), but seeing how ARL doesn't use Foveros Direct, the latency between the two chiplets is going to be much higher, and not at all comparable to the 3D stacking of V-cache.

Broadwell Crystal Well parts were not comparable to 3D stacking of L3 either, in fact they came at extra cost of 2MB of L3 for tags to EDRAM, leaving the chip with less L3 to start with.
Yet they were renowned for years as very good gaming CPUs even with their low 4Ghz clocks and were handily beating the hell out of early 4C Skylake in quite some games before their DDR3 started to drag them down.

I have no idea what and when will be included and why, but such wide claims are intellectually dishonest, when the opposite case can be made: with ARL+ Intel has now lost on-die IMC, average latency has increased and is hurting their performance and "L4" cache setup can only help them regain quite some of that "average" latency.

At a level of power consumption no one cares about or matters. AMD has the most power efficient chips for the vast majority of use cases.
Given that most of the systems spend most of their time idle, that is frankly nonsense claim.

If you're buying a 13900k or 7950x, you almost certainly will be using all those cores, yes. And most of DIY would also be going for gaming as well, which both the 3D and vanilla 7950x are more efficient than the 13900k in

I bought 13900KS and disabled E cores and HT, loving the clocks, loving the efficiency when idle and overall desktop performance and smoothness when loaded.
I am also involved in running multiple 5950x and 7950x systems that have incredible efficiency and load capacity when loaded, but disastrous levels of idle efficiency. 7950x finally has iGPU, so i don't need to add some more idle watts for GPU, great.
 

Tup3x

Golden Member
Dec 31, 2016
1,262
1,390
136
Intel should have skipped this refresh. I fail to see any point in this. Well, at least perf/w graphic will look rather nice when comparing futures stuff to this garbage.
 
  • Like
Reactions: Tlh97 and Thibsie

jpiniero

Lifer
Oct 1, 2010
16,780
7,230
136
Don't know if anyone caught this, but Intel is seemingly confirming that Core (Not Ultra) will be Meteor Lake and not Raptor Lake Refresh-U. Although that then poses the question as to what the difference between the two would be. The Non Ultra having a cheap IGP is the most likely I would think?

Intel should have skipped this refresh. I fail to see any point in this. Well, at least perf/w graphic will look rather nice when comparing futures stuff to this garbage.

Gotta have new products. Even if all Intel can do is minimal improvement.
 

jpiniero

Lifer
Oct 1, 2010
16,780
7,230
136

Raptor Lake Refresh will be the last processor family to use the old brand scheme – officially known as 14th Gen Intel Core processors. The new brand scheme with Intel Core/Core Ultra will begin with Meteor Lake.
 

Hitman928

Diamond Member
Apr 15, 2012
6,694
12,366
136

I read that to say that RPL-R is the last one to use the old naming scheme. That doesn't exclude it from also using the name naming scheme on mobile. Maybe how you suggest is how it ends up, but I think the wording isn't clear enough to call it confirmation of such.