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

[Tigerick]

Wildcat Lake (WCL) Specs

Intel Wildcat Lake (WCL) is upcoming mobile SoC replacing Raptor Lake-U. 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 Q1 2026.

		Intel Raptor Lake U	Intel Wildcat Lake 15W?	Intel Lunar Lake		Intel Panther Lake 4+0+4
Launch Date		Q1-2024	Q2-2026	Q3-2024		Q1-2026
Model		Intel 150U	Intel Core 7	Core Ultra 7 268V		Core Ultra 7 365
Dies		2	2	2		3
Node		Intel 7 + ?	Intel 18-A + TSMC N6	TSMC N3B + N6		Intel 18-A + Intel 3 + TSMC N6
CPU		2 P-core + 8 E-cores	2 P-core + 4 LP E-cores	4 P-core + 4 LP E-cores		4 P-core + 4 LP E-cores
Threads		12	6	8		8
Max Clock		5.4 GHz	?	5 GHz		4.8 GHz
L3 Cache		12 MB		12 MB		12 MB
TDP		15 - 55 W	15 W ?	17 - 37 W		25 - 55 W
Memory		128-bit LPDDR5-5200	64-bit LPDDR5	128-bit LPDDR5x-8533		128-bit LPDDR5x-7467
Size		96 GB		32 GB		128 GB
Bandwidth				136 GB/s
GPU		Intel Graphics	Intel Graphics	Arc 140V		Intel Graphics
RT		No	No	YES		YES
EU / Xe		96 EU	2 Xe	8 Xe		4 Xe
Max Clock		1.3 GHz	?	2 GHz		2.5 GHz
NPU		GNA 3.0	18 TOPS	48 TOPS		49 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.

 

[MoistOintment]

I predicted PTL is gonna be very expensive. Market proved that It is very expensive, and often even more than faster competition, Halo.

MLID agrees both with me and the market.

The market is not fluff, it is reality.

MLID isn't enough of a source on his own.

Is PTL-H 12Xe expensive? Sure. How expensive is it, relative to the competition? And how does this SoC price difference impact the end product when factoring in the increased cost of Ram and SSD? Is it $50/SoC more than Gorgon Point? $20? $100?

You're working backwards by trying to determine the SoC's price based on the launch prices of ultra-premium thin and lights launching with new MSRP's in a DRAM shortage vs discounted models from last year. You can't isolate the SoC pricing based on these handful of laptop model's launch prices.

When 2026 Gorgon Point and Gorgon Halo laptops launch with updated MSRPs impacted by RAM and SSD price increases, what will your reasoning be then? And will you isolate true competitor laptops? Or will you find some chunky plastic laptop with fat bezels, bad speakers, and poor battery life and compare it to a brand new thin and light constructed with a metal chassis that has a tandem OLED screen and draw price conclusions from that?

Unless you can provide some kind of price comparison, even some napkin math based on (alleged) wafer prices and packaging costs, there's no value in these apple-to-orange price comparisons.

 

[fastandfurious6]

(so many words to say nothing)

omg! the market speaks for itself.

posted many times here:

PTL laptop
price: 1500-2500, avg 2000+
gpu: <RTX2050
cpu: ARL > HX370 > PTL > LNL

Halo 392 laptop
price: ~1500-1800
gpu: =RTX4070
cpu: =9900X (12 Zen5 classic) >>> HX370 > PTL


Halo >>>> PTL

/thread

 

[Covfefe]

omg! the market speaks for itself.

posted many times here:

PTL laptop
price: 1500-2500, avg 2000+
gpu: <RTX2050
cpu: ARL > HX370 > PTL > LNL

Halo 392 laptop
price: ~1500-1800
gpu: =RTX4070
cpu: =9900X (12 Zen5 classic) >>> HX370 > PTL


Halo >>>> PTL

/thread

Gift it a rest already and stop thread crapping. You're making up prices. There are Panther lake laptops below 1500 and you've shared zero Halo laptops below 1800 USD.

 

[Abwx]

....

Gift it a rest already and stop thread crapping. You're making up prices. There are Panther lake laptops below 1500 and you've shared zero Halo laptops below 1800 USD.

The only one below 1500$ is no more for sale, if it ever was, that scream false advertisement.

https://twitter.com/x/status/2019180568022290888

 

[Abwx]

Gift it a rest already and stop thread crapping. You're making up prices. There are Panther lake laptops below 1500 and you've shared zero Halo laptops below 1800 USD.

And another view :

https://twitter.com/x/status/2017585098389164219

 

[poke01]

And another view :

https://twitter.com/x/status/2017585098389164219

Dell is cooked. They are pricing their Pantherlake laptops more than a MacBook Pro while offering less performance.

 

[OneEng2]

I predicted PTL is gonna be very expensive. Market proved that It is very expensive, and often even more than faster competition, Halo.

MLID agrees both with me and the market.

The market is not fluff, it is reality.

MLID isn't enough of a source on his own.

Is PTL-H 12Xe expensive? Sure. How expensive is it, relative to the competition? And how does this SoC price difference impact the end product when factoring in the increased cost of Ram and SSD? Is it $50/SoC more than Gorgon Point? $20? $100?

You're working backwards by trying to determine the SoC's price based on the launch prices of ultra-premium thin and lights launching with new MSRP's in a DRAM shortage vs discounted models from last year. You can't isolate the SoC pricing based on these handful of laptop model's launch prices.

When 2026 Gorgon Point and Gorgon Halo laptops launch with updated MSRPs impacted by RAM and SSD price increases, what will your reasoning be then? And will you isolate true competitor laptops? Or will you find some chunky plastic laptop with fat bezels, bad speakers, and poor battery life and compare it to a brand new thin and light constructed with a metal chassis that has a tandem OLED screen and draw price conclusions from that?

Unless you can provide some kind of price comparison, even some napkin math based on (alleged) wafer prices and packaging costs, there's no value in these apple-to-orange price comparisons.

What is absolutely true (and documented) is that 18A development cost Intel > $20B. That's about what a Ford class aircraft carrier costs.

It isn't crazy to think that at the very low volumes Intel is putting through 18A that each wafer is quite expensive.

I agree that product pricing alone is a hard number to use to get down to the CPU cost.

I contend that it doesn't actually matter. What matters is if Intel can make money selling them. In the past, Intel has purchased market share when they wanted (they had LOTS of money) and used this tool like a ball bat on its competitors.

Intel isn't going to be able to do this anymore IMO. It will take a few YEARS to see if 18A paid off.

Right now, fastandfurious6 appears to be right. What you can buy RIGHT NOW with PTL isn't cost competitive with other options.

Will others have to increase prices? Will PTL options go down in prices? Did Intel actually MAKE money selling PTL?

We will see.

 

[OneEng2]

Dell is cooked. They are pricing their Pantherlake laptops more than a MacBook Pro while offering less performance.

That's OK right? After all Dell is considered at the same quality level as a MacBook Pro right .

I have always marveled at Apple's ability to sell at a premium.

 

[poke01]

After all Dell is considered at the same quality level as a MacBook Pro right

Yeah the XPS is good. But they are selling the 8 core Panther Lake for $1599. It should be a 358H 16 core for that price.

 

[Abwx]

Dell is cooked. They are pricing their Pantherlake laptops more than a MacBook Pro while offering less performance.

They are in for the big margins, beside bread and butter are made with early adopters.

 

[desrever]

LOL, obviously price doesn't matter. According to people here and reviewers, only battery life when doing basically nothing and GPU performance below 25W matter.

 

[Doug S]

Same for power usage. Just reset the counter and grab the average package power at the end of the run. It's measured the same way for each system. Works well for comparison purposes if not for absolute accuracy.

No, its really not measured the same way. Maybe if you limit yourself to x86 it is, though even there how you gonna tell me package power is measuring EXACTLY the same thing when you have a CPU that's just a CPU, versus a CPU that includes an iGPU which is the default case in most laptops now. Apple has stuff on chip that's found on the PC motherboard on their SoC, plus stuff that's not even on the board like the SSD controller. So you better make SURE there is no background activity on the SSD during a run or you're overmeasuring power on the Mac.

Finally, for most non hybrid systems you can lock down the max frequency, set it below your cooling system limits and it will sit right there.

Also I should add, that a lot of systems allow for locking max frequency in the BIOS.

That's great, for the systems that allow it. If you want to compare against those that don't?

It's really not as impossible as you make it sound. But yes, it's easier to throw your hands up in the air and say "It's impossible" than to just do it and give up.

Otherwise we argue about much less rigorously tested data, leaks, "I saw this here," etc...

Like I said, I've being doing these types of community tests for over 30 years and have learned quite a bit.

So why you are complaining about the lack of benchmarks that meet your criteria? If you have 30 years of experience in this and know exactly what you want and believe you know how to get it? Why not create this great benchmark yourself?

 

[MoistOintment]

And another view :

https://twitter.com/x/status/2017585098389164219

What exactly does this prove?


1) We already know PTL-H 12Xe is more expensive than LNL. That's obvious

2) We know that existing LNL inventory is very cost competitive because the DRAM was already added to the package months ago

3) The XPS 14 is the replacement to the Dell Pro 14 Premium. Which retails for $1950 on Dell's website with a 236V + 512GB of RAM. That model isn't included in those screenshots, just Dell 14 Premiums.

4) A Dell 14 Premium (Lower tier product than XPS 14) with 255H, 32GB, 1TB with 3.2K display was $2400, on sale for $2000. PTL-H Xe12 in XPS14 with same RAM and storage with 2K display launched at $2049

5) That same Dell 14 Premium with the 255H is over $800 cheaper if you get 2K display, 16GB of RAM, and 512GB SSD

6) Dell Plus 256V + 1TB + 2.5K screen is great value... but it's in a Dell 14 Plus, which is the successor to the Inspiron. XPS being a lot more money than an Inspiron is obvious.

TL;DR According to these screenshots and Dell's website, PTL-H 12Xe is cost competitive to ARL-H if you get the same RAM, Storage, and screen on a Dell Pro 14 Premium vs a Dell XPS 14

What is absolutely true (and documented) is that 18A development cost Intel > $20B. That's about what a Ford class aircraft carrier costs.

It isn't crazy to think that at the very low volumes Intel is putting through 18A that each wafer is quite expensive.

Entirely dependent on the amortization period, which for Intel is what? 7 years? 10 years?

 

[MoistOintment]

LOL, obviously price doesn't matter. According to people here and reviewers, only battery life when doing basically nothing and GPU performance below 25W matter.

Battery life when doing typical office work and home use is a significant demand in the market. Price sensitive customers shopping for e-waste at walmart do just fine with RPL-U, or MTL-U, or Kracken Point. Not like they'd notice a difference. Otherwise they wouldn't have bought e-junk.

It's this forum that over-represents desktop replacements that provide performance at the expense of all else.

 

[Hulk]

No, its really not measured the same way. Maybe if you limit yourself to x86 it is, though even there how you gonna tell me package power is measuring EXACTLY the same thing when you have a CPU that's just a CPU, versus a CPU that includes an iGPU which is the default case in most laptops now. Apple has stuff on chip that's found on the PC motherboard on their SoC, plus stuff that's not even on the board like the SSD controller. So you better make SURE there is no background activity on the SSD during a run or you're overmeasuring power on the Mac.




That's great, for the systems that allow it. If you want to compare against those that don't?



So why you are complaining about the lack of benchmarks that meet your criteria? If you have 30 years of experience in this and know exactly what you want and believe you know how to get it? Why not create this great benchmark yourself?

Thanks for responding. First of all, regarding package power, again, this isn't a moonshot, we're just getting some relative data to compare processors. Some data that is at least gathered in the same manner is more useful than none, right?

As for CPUs, you are correct. I'm only interested in comparing x86 architectures as they are the ones that run Windows and what we are discussing in this thread.

I guess I'm complaining for the same reason that you're complaining about my complaining because that's what we do in these threads, right?

Also, I don't get paid to write benchmark programs for a living. But last I heard I am allowed to comment on what is currently available. Sorry but you don't get to shut me down. I still get to have an opinion and so do you!

But I'm pretty sure I have answered all of your questions and my point still stands that more data is better than less when comparing CPUs. In addition, I still believe that reviewers could just put a little bit more work into obtaining these metrics of performance.

 

[Hulk]

No, its really not measured the same way. Maybe if you limit yourself to x86 it is, though even there how you gonna tell me package power is measuring EXACTLY the same thing when you have a CPU that's just a CPU, versus a CPU that includes an iGPU which is the default case in most laptops now. Apple has stuff on chip that's found on the PC motherboard on their SoC, plus stuff that's not even on the board like the SSD controller. So you better make SURE there is no background activity on the SSD during a run or you're overmeasuring power on the Mac.




That's great, for the systems that allow it. If you want to compare against those that don't?



So why you are complaining about the lack of benchmarks that meet your criteria? If you have 30 years of experience in this and know exactly what you want and believe you know how to get it? Why not create this great benchmark yourself?

Oh, I forgot to mention that there is another way to benchmark hybrid CPUs that I discovered. If you turn off all of the pecores except one and down clock that last one to 900 MHz when you run a single core or single thread benchmark, the thread director is usually smart enough to test one of the more performant cores, which will not be that 900 MHz p core but will be an e core.

It can be kind of fun to think your way around the inherent limitations of locking things down and figuring out ways to isolate those cores. Like the one I just described!

 

[fastandfurious6]

Right now, fastandfurious6 appears to be right. What you can buy RIGHT NOW with PTL isn't cost competitive with other options.

Hallelujah. When was I ever wrong here? So much effort to 'prove me wrong'...

Prophecy #232: PTL price won't drop before zen6 medusa release. medusa will wipe both PTL and NVL out.

 

[DavidC1]

Also, I don't get paid to write benchmark programs for a living. But last I heard I am allowed to comment on what is currently available. Sorry but you don't get to shut me down. I still get to have an opinion and so do you!

It's only possible to have a fair comparison in a desktop. So we'll never get proper comparisons of Meteorlake, Lunarlake, and Pantherlake.

No, its really not measured the same way. Maybe if you limit yourself to x86 it is, though even there how you gonna tell me package power is measuring EXACTLY the same thing when you have a CPU that's just a CPU, versus a CPU that includes an iGPU which is the default case in most laptops now. Apple has stuff on chip that's found on the PC motherboard on their SoC, plus stuff that's not even on the board like the SSD controller. So you better make SURE there is no background activity on the SSD during a run or you're overmeasuring power on the Mac.

It's pointless arguing about this. Fair comparison is not possible on a laptop. We established looking at benchmarks for decades that you need everything to be the same, including down to the weird versions on the OS and on every program.

Hallelujah. When was I ever wrong here? So much effort to 'prove me wrong'...

Oh boy. You really believe this?

What exactly does this prove?

You cannot change the mind of a person that is has their mind set before even going into the argument. The whole point of them arguing is to change yours, in any way possible. But I know you know this so.

 

[DavidC1]

CPU prices aren't always determined by binning, or defects

Sometimes, a perfectly functional CPU has it's cores disabled and clocks reduced and sold at lower prices.

Why? Because it would be financial suicide to sell every CPU at the highest price. It would also be financial suicide to sell it at the lowest price. So, you disable few cores, and sell it as a lower SKU, because certain people are only willing to pay a certain price. It's not a "waste", because otherwise they wouldn't buy it and your revenue is a big fat zero.

I read about when one SKU had unexpected high demand. Intel had to disable many more that could have sold as higher end to fill the missing supply. This is a good indicator that production prices have and are a fraction of the selling price, even before taking into account 3rd party markup(such as to Newegg).

 

[fastandfurious6]

You cannot change the mind of a person that is has their mind set before even going into the argument. The whole point of them arguing is to change yours, in any way possible. But I know you know this so.

I don't write opinions, I write facts and speculation based on facts. Just because many want to prove me wrong doesn't mean I'm wrong. Give a detailed example where I was wrong.

 

[Khato]

I don't write opinions, I write facts and speculation based on facts. Just because many want to prove me wrong doesn't mean I'm wrong. Give a detailed example where I was wrong.

Let's see, despite having been proven wrong in post 24043 - https://forums.anandtech.com/thread...-wcl-discussion-threads.2606448/post-41566289 - about Asus' Strix Halo 392 laptop having an $1800 MSRP according to Asus' own website we continue to have the thread polluted with the below.

omg! the market speaks for itself.

posted many times here:

PTL laptop
price: 1500-2500, avg 2000+
gpu: <RTX2050
cpu: ARL > HX370 > PTL > LNL

Halo 392 laptop
price: ~1500-1800
gpu: =RTX4070
cpu: =9900X (12 Zen5 classic) >>> HX370 > PTL


Halo >>>> PTL

/thread

Show me a Strix Halo 392 laptop that I can buy right now for $1500.

No one disagrees with you that the currently available PTL laptops are expensive. Guess what, the currently available Strix Laptops are equally or more expensive and are not a newly launched product. I also don't think that anyone disagrees that Intel is going to charge more for the 12Xe PTL parts than they are the 4Xe PTL parts. The only point of disagreement seems to be with respect to future pricing of PTL laptops. I see no evidence which suggests that PTL laptops won't come down in price in the exact same fashion as every prior generation has. I don't recall seeing a rational explanation as to why that won't happen?

 

[jpiniero]

Yer buying Strix Halo for the Ayy Eye performance and not gaming.

 

[Hulk]

It's only possible to have a fair comparison in a desktop. So we'll never get proper comparisons of Meteorlake, Lunarlake, and Pantherlake.

What makes these forums "unfun" is when someone has to argue when someone comments, "wow, the sky is really blue today" and someone else, for some reason has to respond, "that's not really blue, it's more bluish, magenta, violet, purple, it is objectively wrong to call a color blue when if you analyze the spectrum of the emitted light, it will not be monochromatic light in the center of spectrum scientists have established as being "blue," therefore your statement is facturally incorrect and without millions of dollars of equipment it is impossible to validate the exact composition of the emitted spectrum by percentage therefor NO comment on color can be made." Then 20 senseless back and forth posts over what constitutes a blue sky.

You can run ST benchmarks on mobile systems as they are almost never power or thermally limited for ST. They hold the max single thread frequency quite nicely. This WILL provide some insight as to "throughput/GHz." Not IPC. This throughput calculation takes all of the factors into account. It's just, what can you do with this much clock? You'll get one number for desktop and another for mobile for a variety of reasons. For example, for CB R26 ST my HX370 holds 5GHz and does 129 points/GHz, while my 9950X at 5GHz does 137points/GHz. Same core architecture and basically the same result as we would expect with the mobile being about 6% less efficient likely due to the slower memory subsystem.

Someone with a Panther Lake could run the ST CB2026 and I guarantee it will hold max frequency. Divide the score by the GHz and boom! You have a comparison between those two specific processors in those two specific laptops. Does it tell us the spectrum of light exactly in the sky? No. But it does give us more insight than the raw scores.

With the data I just provided we know a bit more about how Zen 5 compares in my Asus ProartPX13 as opposed to my desktop 9950X. I like this information. If you don't, that's absolutely 100% fine!

For MT things get a lot more gray as you run into throttling set by manufacturers of each laptop subjectively set based on their testing and comfort levels with heat, power draw, and noise. Those more often than not cannot be easily overidden due to most BIOS parameters being locked in mobile. In addition most benchmarks don't scale linearly with number of cores, making comparisons even more problematic. That is definitely true. But can get isolate these squirmy buggers, even in mobile, with ST benches.

In hybrid systems you can always get a ST P-Core result at max clock. E cores are harder to lock down but if you are clever you can zone in on a good estimate by shutting down P cores and/or using Process Lasso and other methods. When I had Raptors and Alders I got pretty good at it.

All I was saying to to include a bit more data system data with HWinfo with the benchmark result.

I don't need to analyze the spectrum of sunlight reflecting off of the atmosphere to know when it is more or less "blue."

 

[fastandfurious6]

Give a detailed example where I was wrong.

Let's see, despite having been proven wrong in post 24043 - https://forums.anandtech.com/thread...-wcl-discussion-threads.2606448/post-41566289 - about Asus' Strix Halo 392 laptop having an $1800 MSRP according to Asus' own website we continue to have the thread polluted with the below.

Show me a Strix Halo 392 laptop that I can buy right now for $1500.

1. I based my post on this source, posted. It says ASUS priced its TUF HALO 392 32gb laptop at around $1,435. It's not available in market yet so we don't know.


2. What is available is a full ASUS ROG HALO 395+
literal 9950X 16 full cores + RTX4070 level gpu - at $2,169


https://www.newegg.com/asus-rog-flo...-wqxga-32gb-memory-1-tb-ssd/p/N82E16834236572 (and 128GB version for $2799, actually pretty good given it can run huge LLMs)




100% blows apart any PTL laptop in every domain.

Reasonably, ASUS TUF (budget gaming product line, not premium) and HALO 392 (12 full cores 40CU) instead of 395 (16 full cores 40CU) = around the $1500 mark


Find a PTL laptop where it beats the ROG HALO 395 $2,169 in price/performance 😉

Verdict: not wrong at all. Find a real example where I was actually wrong!

 

[Khato]

1. I based my post on this source, posted. It says ASUS priced its HALO 392 32gb laptop at around $1,435. It's not available in market yet so we don't know.

That source is not listing a USD price, it's listing a CNY price. In case you're not aware, many things are cheaper in China. Asus' US website lists a price of $1799 for the laptop, so I'm not sure why there's any question as to the MSRP of the laptop for the US market. Here's the link again for anyone interested - https://www.asus.com/us/laptops/for-gaming/tuf-gaming/asus-tuf-gaming-a14-2026-fa401ea/techspec/