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.

 

[SiliconFly]

.

 

[DZero]

True.

The fantastic thing about Apple's vertical integration is they control every aspect. If Apple wants to switch from ARM to RISC-V, Apple users won't even feel the change. Apple Silicon is good not just because it's ARM, it'll be good even if it's RISC-V or even Apple x86. Thats Apple.

And Microsoft can't do it the same way Apple does. Windows is an open system and there are way too many x86 software dependencies (mostly by 3rd party developers). The entire world is not going to port all their software to ARM just because Microsoft says so. And many of 'em won't even run peoperly on emulation. WoA is a joke.

With that Windows has the days counted since x86 is starting to be defeated by Apple big time.

 

[511]

With that Windows has the days counted since x86 is starting to be defeated by Apple big time.

Of course not it is a double edged sword why would i buy a mac if i knew the next second a support for a feature i use will be killed Enterprises don't like this and people saying x86 is dead is not true

 

[DZero]

Of course not it is a double edged sword why would i buy a mac if i knew the next second a support for a feature i use will be killed Enterprises don't like this and people saying x86 is dead is not true

But for the consumer will start to leave Windows since Apple is starting to mark distances from Windows, and even emulating the x86 products is no problem.

for now the marketshare indicates 15%. Apple is hold back due the high cost, but in the US at least won't be the case for too long.

 

[SiliconFly]

.

 

[511]

But for the consumer will start to leave Windows since Apple is starting to mark distances from Windows, and even emulating the x86 products is no problem.

for now the marketshare indicates 15%. Apple is hold back due the high cost, but in the US at least won't be the case for too long.

View attachment 113770

Linux and chrome os also runs on x86 🫠🫠

 

[poke01]

Of course not it is a double edged sword why would i buy a mac if i knew the next second a support for a feature i use will be killed Enterprises don't like this and people saying x86 is dead is not true

Apple won’t move on from ARM ISA anytime soon.

x86 will be around for a long time due to enterprise, gaming and desktop. What ARM offers is an alternative for in the laptop space. In five years Apple created a sizeable Apple Mac ARM ecosystem.

With regards to windows ARM gaining marketshare it all depends on Gen 3 Cores from Qualcomm and Nvidias ARM SoC.

 

[Nothingness]

Enterprises don't like this and people saying x86 is dead is not true

Many companies do support Mac for their employees. And it works.

 

[techjunkie123]

Many companies do support Mac for their employees. And it works.

I'm an academic and most of my peers have macs for their personal systems. We have linux workstations for compute and older windows workstations to control various instruments.

I'm tempted to move to Apple myself for my laptop. I've had too many problems with my premium Thinkpad laptop to justify it's high price and short battery life (Rembrandt + OLED). The only thing holding me back at this point is my lack of objectivity and the fact that I don't own any other Apple products.

If apple made storage upgradeable on their macs, I'd switch over right now.

 

[Nothingness]

I'm an academic and most of my peers have macs for their personal systems. We have linux workstations for compute and older windows workstations to control various instruments.

I'm tempted to move to Apple myself for my laptop. I've had too many problems with my premium Thinkpad laptop to justify it's high price and short battery life (Rembrandt + OLED). The only thing holding me back at this point is my lack of objectivity and the fact that I don't own any other Apple products.

If apple made storage upgradeable on their macs, I'd switch over right now.

I have switched from a Lenovo P1 running Linux to a Mac Book Pro M1 Max. I'm doing development on it and don't even run a Linux VM, just macOS with Brew packages. I won't look back. For a personal machine, I would likely stick to an x86 PC for gaming; as far as I'm concerned games are the only advantage x86 still has.

BTW I also know companies that don't support Mac. Usually they don't support Linux either. They just want a single platform: Windows and x86.

This is getting off-topic, my point only was that many companies do support Mac. Which doesn't mean x86 is going anywhere (no matter how much I'd like to see it disappear).

 

[511]

Many companies do support Mac for their employees. And it works.

I also work on Macs btw but windows install base on enterprise is huge vs Macs

Apple won’t move on from ARM ISA anytime soon.

Never said they would

 

[SiliconFly]

.

 

[SiliconFly]

.

 

[Raqia]

True.

The fantastic thing about Apple's vertical integration is they control every aspect. If Apple wants to switch from ARM to RISC-V, Apple users won't even feel the change. Apple Silicon is good not just because it's ARM, it'll be good even if it's RISC-V or even Apple x86. Thats Apple.

And Microsoft can't do it the same way Apple does. Windows is an open system and there are way too many x86 software dependencies (mostly by 3rd party developers). The entire world is not going to port all their software to ARM just because Microsoft says so. And many of 'em won't even run peoperly on emulation. WoA is a joke.

In other words, a "consumer electronics" experience: subscription, migration and depreciation by corporate fiat.

The legacy, unsupported, binary blob type dependency in WoA is cleverly addressed with the ARM64EC ABI. I think the cadence for WoA is picking up and will hit critical mass shortly with another potent hardware cycle coming (Qualcomm's first clean sheet client design mostly unimpeded by ARM licensing shenanigans) and concomitant software adoption by big vendors as well as emulation improvements in the form of AVX1/2 support.

 

[SiliconFly]

.

 

[OneEng2]

Many companies do support Mac for their employees. And it works.

Ok, all this talk of x86 death is meaningless without some hard numbers .... which none of you have provided. Anecdotal evidence is silly. I don't understand for the life of me why people do it.

"George Burns smoked his entire live and lived to 100 years old, therefore smoking is good for you".

"I use a mac at work therefore x86 is doomed"

Here is a fact. 98% of corporate computers world wide use x86. The market is price conscious, and risk adverse. The insanely high amount of dependent applications on x86 presents a virtually insurmountable barrier to any other ISA. Shy of a federal LAW (which will never happen), x86 will remain dominant for at least another 10 years. My guess is 20 ..... or more.

as far as I'm concerned games are the only advantage x86 still has.

... and you represent how much of the market again?

Intel is in trouble for sure, but it is about their financial position and business model. It has NOTHING to do with ARM or RISC or any other boogie man out there.

Despite all the gloom and doom about Intel, they have 75% of the x86 market. People need to think about the logistics here a little. There is no way, what-so-ever, that Intel could be allowed to fold up shop. Be purchased? Only by another American company.... that much is for sure.

Besides, so many people here focus on gaming performance. Yes, AMD has a large advantage in this space .... but it is such a sliver of the x86 market! From a performance standpoint ARL is a strong competitor to Zen 5.

My gripe about ARL has nothing to do with doom and gloom of its performance, or how effective x86 will be in the future, or how well Intel will do with the Panther Lake follow on product.

My concern is that Intel is paying so much for a premium node at TSMC for ARL and LNL while paying out the nose for 18A ramp up at the same time. I am a little concerned that ARL doesn't best Zen 5 across the board while being produced on a higher density, lower power node .... but seriously, considering the Awful latency figures the processor has, it is doing an amazing job. I think people really underestimate how much performance Intel's current architecture is going to be able to unleash with the next tweak or two..... and this is without the improved density that 18A with BSPDN will provide.

Now, don't get me too wrong here. I am an AMD fan and have been for decades. I love me a long lasting socket with on-going cheep upgrades over many years on each platform (which Intel has denied me for that same number of decades). But the fact remains, I (and anyone like me) am a minority as hardly ANYONE builds their own computer anymore. Shoot, I might just bite the bullet and buy a laptop next time myself. If you don't game, it gets harder and harder to justify a desktop.

 

[poke01]

They can and they might at some point. It's not that difficult for them. Considering they aren't getting much gains out of ARM, RISC-V is a better choice for a do over.

Not happening for another 15-20 years

 

[poke01]

Ok, all this talk of x86 death is meaningless without some hard numbers .... which none of you have provided. Anecdotal evidence is silly. I don't understand for the life of me why people do it.

Almost of Top 100 US companies provide an employee with either a ARM MacBook or Windows laptop.

Amazon, Nvidia, Google, Microsoft, Netflix, Adobe to list some of top US companies.

Heck even the medium sized company in Australia which is not even on the AUS stock exchange provides MacBooks to some devs and executives.

Yes, AMD has a large advantage in this space .... but it is such a sliver of the x86 market! From a performance standpoint ARL is a strong competitor to Zen 5.

AMD is not huge in the client space, especially OEM desktops and laptops. It’s been 6 months since Zen5 mobile launched and the majority of laptops are still from ASUS.


Some here need to get rid of the x86 only mindset. By the end of this decade the laptop landscape will change drastically. Qualcomm launched with a different variety of OEMs on day 1!!

 

[Hulk]

This is why E-cores are useless to me. They are missing so many features and cause scheduling problems. Not worth it. And now that Intel has access to smaller nodes (theirs or TSMC) whey SHOULD be able to create powerful, efficient low power CPUs with all the features.

How come the logical cores don't cause scheduleing issues with AMD processors?

Follow me here.

We have a thread that needs to go to a physical core with Zen 5 but instead is incorrectly assigned to a logical core. This would seem to create a more catastrophic decrease in performance than a thread intended for Lion Cove going to Skymont. My reasoning is that the performance delta between Zen 5 physical and logical cores is much greater than the performance delta between Lion Cove and Skymont.

While I am obviously a fan of Zen 5, now that they both have two different types of cores that threads need to be appropriately scheduled to, why is thread scheduling more problematic on ARW compared to Zen 5? As noted above incorrect thread scheduling would seem to be more problematic from a performance point-of-view than with ARL given the wide performance delta between Zen 5 logical and physical cores.

 

[Raqia]

So, with ARM64EC ABI, when do you think WoA will replace Winx64? 2026 or 2027?

Would say Windows is its own thing, and for the most part largely similar across the 2 major supported platforms with compatibility (and possibly gaming support) being its key feature. One of Microsoft's goals with NT was cross platform support, and I would view the latest builds of Win 11 which axes support for some legacy x86 CPUs and adds better support for ARM an attempt to unify both the code base as well as the experience:

"We really focused on modernizing this update of Windows 11," said Microsoft Corporate Vice President of Windows and Devices Pavan Davuluri at a technical briefing on Microsoft's campus in mid-April. "We engineered this update of Windows 11 with a real focus on AI inference and taking advantage of the Arm64 instruction set at every layer of the operating system stack. For us, what this meant really was building a new compiler in Windows. We built a new kernel in Windows on top of that compiler. We now have new schedulers in the operating system that take advantage of these new SoC architecture."

Microsoft didn't say whether the updated system components would have user-noticeable benefits for users of current x86 systems, though these updates are likely the reason why the OS has gone from "unsupported" to "unbootable" on some systems with early 64-bit x86 processors.

It’s not “Windows 12”: Microsoft keeps Windows 11 branding despite major changes

24H2 update comes with new compiler, kernel, and scheduler, among other changes.

arstechnica.com

There's no very good reason to axe legacy x86 CPUs except to modernize, streamline and unify higher level code base support that now supports ARM instructions. As for the new deficit in hardware support: I would argue supporting old software is vastly more important to most corporations than old hardware as costs in time, money and effort to redevelop and validate old software easily swamps the cost of updating hardware.

So I'd say the version of Windows 11 that supports ARM has already "replaced" the older Windows mostly built around only the X86-64 platform.

 

[Markfw]

How come the logical cores don't cause scheduleing issues with AMD processors?

Follow me here.

We have a thread that needs to go to a physical core with Zen 5 but instead is incorrectly assigned to a logical core. This would seem to create a more catastrophic decrease in performance than a thread intended for Lion Cove going to Skymont. My reasoning is that the performance delta between Zen 5 physical and logical cores is much greater than the performance delta between Lion Cove and Skymont.

While I am obviously a fan of Zen 5, now that they both have two different types of cores that threads need to be appropriately scheduled to, why is thread scheduling more problematic on ARW compared to Zen 5? As noted above incorrect thread scheduling would seem to be more problematic from a performance point-of-view than with ARL given the wide performance delta between Zen 5 logical and physical cores.

Zen5c and Zen5 cores all have the same capabilities totally different than Intels P and E core scenario. If it went to the wrong core it would only run a LITTLE slower, not a lot. You would never notice. (unless you ran a Zen5 only bench with everything else the same., that does not exist most likely)

 

[biostud]

With regards to windows ARM gaining marketshare it all depends on Gen 3 Cores from Qualcomm and Nvidias ARM SoC.

Or more likely how much Microsoft is going to invest in making Windows on Arm a viable alternative.

 

[igor_kavinski]

While I am obviously a fan of Zen 5, now that they both have two different types of cores that threads need to be appropriately scheduled to, why is thread scheduling more problematic on ARW compared to Zen 5? As noted above incorrect thread scheduling would seem to be more problematic from a performance point-of-view than with ARL given the wide performance delta between Zen 5 logical and physical cores.

Context switching latency between a Lion Cove core and a Skymont cluster is much higher. Also, the virtual thread in SMT is just an illusion. To the core, it's just two instruction streams in flight so the core will just crunch through those instructions as quickly as possible because SMT works by boosting resource utilization to the max, ensuring there is no idle time betweens loads, stores etc.

Intel probably has data showing how idle Lion Cove core resources get during certain workloads but of course they won't share that publicly because just like their stupid E-core marketing spiel, this time they are doing the world a "favor" by dropping HT and concentrating on ST performance which by the way still didn't get to be the best in the world. Really looking forward to them releasing a refresh with HT enabled.

 

[adroc_thurston]

Really looking forward to them releasing a refresh with HT enabled.

It's physically not present in ARL.

 

[igor_kavinski]

It's physically not present in ARL.

They better make it happen! I don't care if they have to take their server core and splice it into the consumer design.