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]

.

 

[AcrosTinus]

I know many here aren’t fans here of MILD, but he has commented on the Royal Core cancellation theme. If what he’s saying in the video is legit, it sounded pretty cool:
Dynamic splitting of threads per core. 2x IPC of RPL. Says this was from a face-to-face meeting with a verified Intel employee aware of the project.

AGAIN, Royal Core is not real, it is not a product. You cannot cancel something that never was. This is a normal R&D exercise in the labs. After things are deemed viable or not, aspects are extracted and used in other products.

How can people be this gullible?
He is basically extracting you for views, feeding you stories that meet your interests, he could basically make things up and people will listen to him.

In reality this video if from a guy that is known to fabricate stuff, claiming to allegedly be talking to a unverified Intel source that somehow spilled the beans to a anti Intel youtuber.

No wonder media literacy is low and idiots can influence stock markets. If fairy tails are given the same weight as researched outlets we are doomed, RIP AnandTech. You inspired me to go through electrical engineering much love.

 

[Kocicak]

AGAIN, Royal Core is not real, it is not a product. You cannot cancel something that never was. This is a normal R&D exercise in the labs.

What are you talking about? You consider "a real product" only what has been officially anounced by the manufacturer?

So you REALLY KNOW future planned Intel CPUs are you are sure that Royal core (the 6 core 6-24 thread rumored CPU) has never been seriously planned as one? Are you Intel employee?

May I ask you to explain typology of projects in Intel depending on how seriously is the outcome of the project considered to become a real product in the future and how are these projects being approved, financed and managed?

Thank you.

 

[Magio]

The primary driver for better battery life from those products is their ST power, so I don't understand how LnL will achieve these expectations.

The SoC may provide good idle , good HW accelerated video consumption numbers, but to get these into the real world battery life advantage will take very disciplined platform engineering, and without a program like Evo on steroids , and focusing their marketing on specific devices I don't think there'll be much to show for it. but we'll see.

The key for battery life on LNL (besides SoC and media engine improvements) will be how many loads they'll be able to confine to the Skymont cluster (and of course how efficient that cluster is). LNC seems like a marked improvement in ST efficiency but it still doesn't quite compare to Oryon or especially to Apple Silicon in terms of PPW.

But if you can avoid having the LNC cores turning on as much as possible, and limit them mostly to heavy loads (where, let's be real, battery life is mostly a factor of TDP/battery size), then that's not too big a problem as far as battery life is concerned.

 

[cannedlake240]

Intel has always been a microprocessor-first company

Intel was once a memory company. Then their founding fathers managed to (very successfully) pivot to Intel to the CPU business. I guess Pat, inspired by this past glory and the Nvidia's profits, now wants Intel to become an AI + Foundry company selling CPUs on the side. The problem is, just this week he admitted that Intel isn't competing with the best of Nvidia, Google and AMD anytime soon. Hope Falcon Shores is really promising, otherwise dissolving a design team after half a decade of work on an innovative uarch is just bizarre... The AI hype might end any time now while Client & DC CPUs aren't going anywhere

 

[511]

And Bob Swan should have fixed it. But he didn't care about the future of Intel foundries. Instead, the knucklehead was more interested in Intel stock prices buy backs.

Agreed

Like I mentioned before, both the biggest knuckleheads in Intel's history, Brian Krzanich & Bob Swan, worked real hard to run Intel into the ground and they succeeded. Now, Pat Gelsinger is trying out a lot of new tricks, but unfortunately, the results have been subpar.

At least Pat is trying he is not being successful

For example:
Intel 4/3 don't have enough capacity.
Intel 20A is delayed.
They're bleeding server market share.

Intel 3 is a very good node it's just they lack machines it's PPW is around N3B/E and density similar to N4 but the lack of capacity is an issue
20A is like Intel 4 a public beta node for few sku for learnings

Their desktop client product quality is subpar.
Their mobile client MTL is a colossal failure (imho).

Meteor lake is a failure it sold 15 million+ units 😂
Tiger lake was a success sold 100Million+

This isn't the Intel I knew. If Pat doesn't pull a rabbit out of his hat real soon, he too might join the list as the third biggest knucklehead in Intel's history following Brian Krzanich & Bob Swan.

His rabbit is named 18A/PTL/CLF

Imagine, we can call them *3 idiots*!

4 we already have paul simply refused Apple iphone SOC

 

[Thunder 57]

And Bob Swan should have fixed it. But he didn't care about the future of Intel foundries. Instead, the knucklehead was more interested in Intel stock prices buy backs.

Like I mentioned before, both the biggest knuckleheads in Intel's history, Brian Krzanich & Bob Swan, worked real hard to run Intel into the ground and they succeeded. Now, Pat Gelsinger is trying out a lot of new tricks, but unfortunately, the results have been subpar.

For example:
Intel 4/3 don't have enough capacity.
Intel 20A is delayed.
Intel 18A not enough customers.
They're bleeding server market share.
Their desktop client product quality is subpar.
Their mobile client MTL is a colossal failure (imho).

This isn't the Intel I knew. If Pat doesn't pull a rabbit out of his hat real soon, he too might join the list as the third biggest knucklehead in Intel's history following Brian Krzanich & Bob Swan.

Imagine, we can call them *3 idiots*!

That's the problem with an interim CEO.

Let's play a game, I'm the golden trickster fish and you're Pat Lucky. You found me while fishing, and you get one wish to save Intel from the ones below:

• Intel gets best SoC of 2025 with Lunar Lake and Arrow Lake wins vs. Zen 5 in performance and efficiency (ARL refresh included)
• Intel matches or beats TSMC with their fifth node in 4 years, and is set to win foundry contracts for short and mid term

Why am I the trickster golden fish, you ask? Because when you pick one, the other is automatically excluded. In a way this matches reality, it would take a miracle for both of the above to happen at the same time. So which one do you pick to save Intel? 🐡

I'd say A is more likely but B would be far better for Intel in the long run.

 

[AcrosTinus]

What are you talking about? You consider "a real product" only what has been officially anounced by the manufacturer?

So you REALLY KNOW future planned Intel CPUs are you are sure that Royal core (the 6 core 6-24 thread rumored CPU) has never been seriously planned as one? Are you Intel employee?

May I ask you to explain typology of projects in Intel depending on how seriously is the outcome of the project considered to become a real product in the future and how are these projects being approved, financed and managed?

Thank you.

We have no real insight, what MLID is offering is conjecture and is worth nothing, he has no ethics.
And the first sentence is on the money, a real product is what is announced by the manufacturer and can be bought by consumers or businesses.
Everything else is internal development that is not meant for the public, meaning when an R&D exercise does not pan out you cannot speak about a cancelled product, closer would be a potential idea that was deemed not viable. Even then the confidence in such information should be extremely low.

But in the age of fake news and misinformation you decide your media diet.... good luck.

 

[cannedlake240]

already have paul simply refused

They somehow had a relatively smooth transition to 22nm, but the signs of trouble brewing were already there, it's said. This new GPU/AI push better not be a repeat of the mobile debacle they had a decade ago

 

[Kocicak]

what MLID is offering is conjecture and is worth nothing, he has no ethics.

A person delivering a message and the content of the message and two different things and even a pathological liar sometimes tells the truth.

So you automatically presuming that the information coming from the source you do not like is false is simply an incorrect inference.

 

[majord]

The key for battery life on LNL (besides SoC and media engine improvements) will be how many loads they'll be able to confine to the Skymont cluster (and of course how efficient that cluster is). LNC seems like a marked improvement in ST efficiency but it still doesn't quite compare to Oryon or especially to Apple Silicon in terms of PPW.

But if you can avoid having the LNC cores turning on as much as possible, and limit them mostly to heavy loads (where, let's be real, battery life is mostly a factor of TDP/battery size), then that's not too big a problem as far as battery life is concerned.

I agree, but if you can confine a decent amount to E cores, then begs the questions why 4+4? , why not 2+6 or something. This is the reason I find big Little architectures..(and that includes AMD APU's C core arrangement ) a bit of an enigma at times . They make sense fundamentally, but how does one decide what balance is appropriate? It's really trickly in the wonderfully unpredictable world of x86 ,. Even if it's not optimal, it won't show in ST MT benchmarks

 

[Magio]

I agree, but if you can confine a decent amount to E cores, then begs the questions why 4+4? , why not 2+6 or something.

You probably need the quad core LNC cluster for gaming, for example. And of course, LNL won't be a gaming powerhouse but it does have a beefy GPU for its TDP so clearly that was part of the focus.

Personally I feel like LNL is a very well balanced chip on paper, looking to strike a balance between efficiency and breadth of viable use cases. Of course, now it has to confirm that "in the field".

 

[511]

Personally I feel like LNL is a very well balanced chip on paper, looking to strike a balance between efficiency and breadth of viable use cases. Of course, now it has to confirm that "in the field".

3 days before the NDA is lifted

 

[SiliconFly]

.

 

[SiliconFly]

.

 

[Thunder 57]

It's even clear from MLID's video where he says he asked keller "how to combine multiple threads to make a single thread to boost IPC". Stupid question. He says keller replied it's actually the opposite which is very true. I'm actually appalled by MLID's ignorance. If he can't grasp such basic concepts, we need to question each & every assessment he makes based on his leaks.

What are you talking about? Reverse Hyperthreading is total a thing! /s

 

[igor_kavinski]

What are you talking about? Reverse Hyperthreading is total a thing! /s

We have enough deep knowledgeable techies on this forum. We should create our own version and patent it and then we can buy AT from Future PLC and do whatever we want with it!

Come on, guys! Let's ASSEMBLE!!!!

 

[coercitiv]

3 days before the NDA is lifted

Oh, nice. Finally something to look forward to!

 

[Magio]

3 days before the NDA is lifted

Well I doubt we'll know *everything* on September 3rd. Mobile platforms can only really be judged by the quality of the designs they'll be integrated in, so it will take until LNL design wins are revealed and released into the hand of reviewers.

But we'll know a lot more as it launches and during IFA in general.

 

[igor_kavinski]

Let's play a game, I'm the golden trickster fish and you're Pat Lucky. You found me while fishing, and you get one wish to save Intel from the ones below:

• Intel gets best SoC of 2025 with Lunar Lake and Arrow Lake wins vs. Zen 5 in performance and efficiency (ARL refresh included)
• Intel matches or beats TSMC with their fifth node in 4 years, and is set to win foundry contracts for short and mid term

Why am I the trickster golden fish, you ask? Because when you pick one, the other is automatically excluded. In a way this matches reality, it would take a miracle for both of the above to happen at the same time. So which one do you pick to save Intel? 🐡

TBH, I don't think any of those will happen.

Lunar Lake will be popular with OEMs the same way MTL was (so nothing out of the ordinary for Intel).

Arrow Lake will sell decently but won't set any performance records. The RAM overclocking ability will satisfy the existing LGA1700 overclockers into upgrading and trying to reach new DDR5 heights.

As for matching TSMC? Maybe but no way large volume customers are gonna bet on them ONLY for their precious product. Their fabs/process may be used as a backup or to try out less ambitious product SKUs to see how well Intel serves them and their needs.

If there is going to be a miracle, it's going to be something secret that Intel has managed to keep hidden from public view so far.

 

[511]

Come on, guys! Let's ASSEMBLE!!!!

ARM assembly or x86 Assembly pick your poison or RISC-V as a hardcore option

 

[511]

Well I doubt we'll know *everything* on September 3rd. Mobile platforms can only really be judged by the quality of the designs they'll be integrated in, so it will take until LNL design wins are revealed and released into the hand of reviewers.

But we'll know a lot more as it launches and during IFA in general.

David Huang already has microbenchmarked everything he will release the data on 3rd

 

[Magio]

As for matching TSMC? Maybe but no way large volume customers are gonna bet on them ONLY for their precious product. Their fabs/process may be used as a backup or to try out less ambitious product SKUs to see how well Intel serves them and their needs.

Capacity will come, or at least the plans for that aren't kept secret:

Intel 18A and Intel 3 should remain attractive nodes for a good while (including their refinements 18A-P, Intel 3T/E/PT), let's not forget that even today the best chips from Qualcomm, AMD and Nvidia are still on TSMC's 4nm processes and those are very lucrative contracts for TSMC. So if they can ramp the capacity on those as planned, there will be contracts to be found aplenty.

If those are rolled out without many issues, despite the initially limited capacity I can only imagine the prospect of uncontested leadership with 14A would attract extremely lucrative contracts. Of course, great nodes would also be a boon for the product side of the company albeit Intel will have to be careful not to withhold too much of their capacity for themselves if they want to attract the biggest guns in the chipmaking industry.

Of course I'm not saying any of that will come to be, I'm no sightseer. But the plan is there and it is extremely promising on paper.

 

[igor_kavinski]

Here's a prediction: Despite lacking hyperthreading, LNL (both P-cores and E-cores) will perform decently in 1080p gaming workloads because Skymont is roughly "Raptor Cove"-powerful and Lion Cove is supposed to be "Raptor Cove++". So any game threads juggling between heterogeneous cores shouldn't result in a wide discrepancy in performance. If this proves to be true and I can find a decent model that I can get with monthly installments, Imma gettin' it!

 

[igor_kavinski]

But the plan is there and it is extremely promising on paper.

Their execution on promised plans so far has been meh. I guess they think their investors are the type who obsess about two birds in the bush rather than the one in their hands.