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.

 

[Io Magnesso]

It seems that there are quite a few SKUs.

 

[MoistOintment]

Question should be, "How is that different from a PC?"

PC has open standards so multiple operating systems exist for it including BSD, Linux, Windows and of course, some people run hacked MacOS on their PCs too.

Contrast that with the closed and shrouded in secrecy Mac architecture where a small group of volunteers have been trying to reverse engineer it and still nowhere close to getting Linux fully supported on it. So the only way to run your application is to either develop it from scratch on MacOS, look for an existing available application outside the Apple Store (mostly fat chance) or use one on the Apple Store (better chance as developers prefer the wider userbase they can reach through that). So while MacOS applications are available outside the Apple Store, there just isn't a wide variety of them available due to the comparatively smaller userbase. Fixing this situation requires them to post the required patches to the Linux kernel for their hardware or document their hardware enough so people can do it themselves. As for Windows, need to wait and see what happens when Nvidia N1X hardware is available with Windows on ARM because that will signal the end of the exclusivity period with Qualcomm.

You're just explaining why Apple is motivated not to enter the DC market. Not why they can't. All of these problems you list are decisions that Apple intentionally made as a client-focused company. If they did choose to enter the DC space, these "issues" could easily be resolved.

 

[Io Magnesso]

Is it MacOS…
Apple free-riding to ecosystems of other companies and other platforms...
Game transplant tool...(Windows) containerized(Linux)
At least Windows and Linux contribute or provide each other while using each other's ecosystem That is, there is a return
But... the above story is one-sided.
It's a very convenient story, I can't complain...

 

[Magio]

Oh that’s more like it. The 10xe core sku is good

The original leaked SKUs are probably the base die configs, and then most likely 4+4+4+10 and 4+4+4+4 will be derived from 4+8+4+12/4+8+4+4 parts with some disabled cores due to defects.

But yeah, PTL will cover a *much* wider range than LNL did. LNL was ultimately just one die config and some down bins. Do wish there'd been a PTL-M to really be a 1 to 1 replacement of LNL but that wasn't to be.

 

[511]

The original leaked SKUs are probably the base die configs, and then most likely 4+4+4+10 and 4+4+4+4 will be derived from 4+8+4+12/4+8+4+4 parts with some disabled cores due to defects.

But yeah, PTL will cover a *much* wider range than LNL did. LNL was ultimately just one die config and some down bins. Do wish there'd been a PTL-M to really be a 1 to 1 replacement of LNL but that wasn't to be.

Don't worry NVL-AX has MoP under consideration to hit higher DDR Speeds than what's possible with PCBs due to signalling issues.

 

[okoroezenwa]

But... the above story is one-sided.
It's a very convenient story, I can't complain...

Also not a meaningful story really

 

[Io Magnesso]

Also not a meaningful story really

Right, As for the thread about the Intel CPU...
Well, I also like Apple's products, but I don't like it when I affirm all their actions.
No matter how good a company the other party is, there is no perfect thing, so we have to give feedback...

 

[poke01]

Well look at this

Core Conversations: Apple M4 vs Lunar Lake/ Arrow Lake

This podcast discusses the differences in performance and abilities of the Apple M4 product line vs Intel Lunar Lake and Arrow Lake processors

www.intel.com

Ngl, they made some great points

 

[511]

Well look at this

Core Conversations: Apple M4 vs Lunar Lake/ Arrow Lake

This podcast discusses the differences in performance and abilities of the Apple M4 product line vs Intel Lunar Lake and Arrow Lake processors

www.intel.com

Ngl, they made some great points

The biggest advantage for Apple is not having to deal with Insane combination of HW like x86 Vendors.

 

[Io Magnesso]

On a platform called x86, it's a shared asset.
Users can choose Intel or AMD as they like, and can be migrated relatively easily.
Well, there are only two companies on the x86 platform, except for Chinese companies...
It's a little lonely... It would be nice if there was a rival who would fight on the same x86 platform, not the rival called ARM.

 

[Io Magnesso]

The biggest advantage for Apple is not having to deal with Insane combination of HW like x86 Vendors.

Well, it's easy if you can manage everything by hand. As expected, the current Intel does not have enough space, but for example, from Intel's official It's the same feeling as Intel's motherboards and NUCs appeared in the past.
It may be easier for consumers to choose as an indicator if they create a PC product group made by Intel.
Or rather, Intel is already NUC Laptop I was releasing that

 

[igor_kavinski]

It's a little lonely... It would be nice if there was a rival who would fight on the same x86 platform, not the rival called ARM.

First x86 SMT4 CPU: https://www.tomshardware.com/pc-com...with-avx-512-and-16-channel-ddr5-5600-support

I hope it's available outside China at some point. Would be fun to benchmark it, even if it's otherwise not very competitive with even 5 year old CPUs from Intel and AMD.

 

[Io Magnesso]

Well, I know that there are several manufacturers in China that have an X86 license.

 

[511]

Well, I know that there are several manufacturers in China that have an X86 license.

It's required by the law they have to share IP if you want to operate business in China.

 

[DavidC1]

The original leaked SKUs are probably the base die configs, and then most likely 4+4+4+10 and 4+4+4+4 will be derived from 4+8+4+12/4+8+4+4 parts with some disabled cores due to defects.

Emphasis on "some". Most are artificially disabled because the market won't accept certain price points. So some have to be cut down so they can sell it at a lower price, or they won't be sold.

There was a story when Intel didn't have enough Celeron chips so they had to disable perfectly working Pentium dies and resold as Celerons(these were when Pentiums were the top dog).

Why not just sell Pentiums at a lower price? Because it will merely lower revenue. Silicon is cheap, losing a customer is not.

Than wtf is Intel doing?????!

I ain’t paying for the top sku for the 12 GPU cores that’s gotta cost a lot of $$$. Lunar lake comes with 7-8 on the usable SKUs. 226V is 7 core and 256V is 8 core.

I can tell that 12Xe3 SKU will be low volum just like the 288V..

You know why they do it right? Because they want to upsell. Not all will fall for it, but some will.

Modern iGPUs if you think about it aren't free. Before with the chipset graphics you could pair the best one with a Celeron. You can't do that anymore. If you look at the price, essentially you are forking $100-200(or more) for the fastest iGPUs. So price adder went from $5 to 10-40x that.

Don't worry NVL-AX has MoP under consideration to hit higher DDR Speeds than what's possible with PCBs due to signalling issues.

I HIGHLY doubt MoP is the only difference between Lunarlake and Pantherlake. There are many small details they haven't said.

I just don't believe Pantherlake will be similar in efficiency as Lunarlake. Forget what they say, even in technical presentations. Look at what they do. Think of why Dell is not having a direct Pantherlake successor to Lunarlake? It should be cheaper, so beneficial for Dell right?

They don't need MoP to have it that low power. Bay Trail did that 11 years ago. Arguably it was even lower power than Lunarlake. My Venue 8 Pro idled at 1.6W with screen on! Youtube browsing with multiple tabs I got 6-7 hours of battery with it's puny 17WHr capacity.

What should have continued is a dedicated low power platform. But "Muh profits", so they don't do this.

Intel is the Microsoft of the hardware world. They are a perfect pair because they make stupid decisions together.

 

[511]

I HIGHLY doubt MoP is the only difference between Lunarlake and Pantherlake. There are many small details they haven't said.

PMIC/MoP with optimized PHY For Power are the only difference between Lunar and Panther Panther is a derivative of Lunar.

I just don't believe Pantherlake will be similar in efficiency as Lunarlake. Forget what they say, even in technical presentations. Look at what they do. Think of why Dell is not having a direct Pantherlake successor to Lunarlake? It should be cheaper, so beneficial for Dell right?

If you are wondering about Dell Leaked Roadmap maybe they updated it later also it may have to do something with the agreement with Intel we want this for at least till X/Y/Z year besides Lunar is an awesome product.

They don't need MoP to have it that low power. Bay Trail did that 11 years ago. Arguably it was even lower power than Lunarlake. My Venue 8 Pro idled at 1.6W with screen on! Youtube browsing with multiple tabs I got 6-7 hours of battery with it's puny 17WHr capacity.

What should have continued is a dedicated low power platform. But "Muh profits", so they don't do this.

Muh Shareholders Feets need to lick em more

Intel is the Microsoft of the hardware world. They are a perfect pair because they make stupid decisions together.

Yeah

 

[Magio]

I just don't believe Pantherlake will be similar in efficiency as Lunarlake. Forget what they say, even in technical presentations. Look at what they do. Think of why Dell is not having a direct Pantherlake successor to Lunarlake? It should be cheaper, so beneficial for Dell right?

Well this is not so straightforward. LNL will be on the 20W TDP "XPS" 13 model (per the leaked roadmap), so the only thing they could have put there would have been a PTL-U part but then it would have been worse on the iGPU. They'll have PTL on the 40W "XPS" 14, and if that next "XPS" 13 was actually targeting 30W it would probably be PTL too*. (Side note, how dumb are Dell for killing the XPS brand...)

This tells us that PTL has nothing for LNL's exact segment, but it doesn't tell us PTL is less efficient. It's just a gap in the PTL line-up.

Now do I think PTL is as efficient as LNL in every way? No. But do I think the ways it will be more efficient (I believe 18A will be a measurably better node than N3B, arch refinements, media engine refinements, ...) will make it overall as efficient or more? I do actually believe that. But we won't know for sure until it launches of course.

*I'll also add that LNL at the 20W range ended up being a bit less good than once believed, so since that roadmap is quite old it's not impossible Dell *did* end up upping the Wattage of that next gen 13 model, and if they did they might in fact have moved away from LNL at some point.

 

[Io Magnesso]

Well this is not so straightforward. LNL will be on the 20W TDP "XPS" 13 model (per the leaked roadmap), so the only thing they could have put there would have been a PTL-U part but then it would have been worse on the iGPU. They'll have PTL on the 40W "XPS" 14, and if that next "XPS" 13 was actually targeting 30W it would probably be PTL too. (Side note, how dumb are Dell for killing the XPS brand...)

This tells us that PTL has nothing for LNL's exact segment, but it doesn't tell us PTL is less efficient. It's just a gap in the PTL line-up.

Now do I think PTL is as efficient as LNL in every way? No. But do I think the ways it will be more efficient (I believe 18A will be a measurably better node than N3B, arch refinements, media engine refinements, ...) will make it overall as efficient or more? I do actually believe that. But we won't know for sure until it launches of course.

Certainly, the now-deceased XPS brand must have had a variant with a slightly looser power limit. The one that can be loaded with a larger CPU
There are also models that can be installed in dGPU.

 

[Io Magnesso]

And in the present age, there is a convenient thing called cTDP.

 

[Doug S]

The biggest advantage for Apple is not having to deal with Insane combination of HW like x86 Vendors.

The hardware combinations are a problem for Microsoft & Linux not for Intel & AMD.

 

[Io Magnesso]

The hardware combinations are a problem for Microsoft & Linux not for Intel & AMD.

Well, it's not just Intel or AMD...
Peripheral equipment manufacturers...

 

[Doug S]

Well, it's not just Intel or AMD...
Peripheral equipment manufacturers...

Its a software problem for them too. They have to provide drivers for the OSes they want to support. I'm not aware of any cases where say two PCIe cards conflict with each other or two USB devices conflict with each other, it isn't like the old days where they'd have IRQs that would step on one another or the like.

 

[Io Magnesso]

Its a software problem for them too. They have to provide drivers for the OSes they want to support. I'm not aware of any cases where say two PCIe cards conflict with each other or two USB devices conflict with each other, it isn't like the old days where they'd have IRQs that would step on one another or the like.

Well Linux(UNIX) In the case of It's not perfect
It is a blessing because there are some private collaborators and collaborators from other companies.

 

[poke01]

The x86 platform supports a lot of hardware, has compatibility etc goes out the window with Windows on ARM. As it is now, Qualcomm will release a much better CPU later this year and all those points become invalid.

 

[poke01]

There is a reason why Intel made fun of Qualcomm last time. They are a threat to Intel, not Apple.