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.

 

[igor_kavinski]

The only guy capable to do turn around of that scale is dead unfortunately

No one is indispensable. In every day and age, there is always a worthy successor. Whether the people in charge really want that person to lead the effort is the real question. Here's the thing: a truly worthy person will tell the Intel board things they DO NOT want to hear. Hence, they will keep appointing jackasses. But maybe this time they should go with an external jackass. Anyone other than another Intel jackass.

 

[511]

Yes, all the fabs other than TSMC that are doing just fine and dandy are committing suicide every single day

Samsung is in a worse condition than Intel in terms of nodes and we don't have leading edge logic maker besides those 3 for Samsung they have the Korean government support and for TSMC they have Taiwan's government support and US government doesn't have much faith in Intel I guess

 

[SiliconFly]

.

 

[Thunder 57]

Maybe we should start a CEO thread. Pat, Lisa, Jensen, fight! Just sayin'

 

[igor_kavinski]

So, by your definition, Lisa is a bigger jackass.

Nah. She's enjoying the show. She will just sack the graphics guy if his plan fails because this is not her plan. She goes after the big stuff. Like getting people from diverse backgrounds to innovate together. She has her plate full at the moment and frankly, even I wouldn't want to be doing anything consumer Radeon related unless I had unlimited budget and complete immunity from sacking over moronic mistakes. The 9070-do-not-launch "apparent" fiasco is just a last ditch effort to make the most of a bad situation (not having any halo product when the market is lusting after limited RTX 5000 series supplies).

Why are you spewing so much venom against Intel in Intel thread?

Didn't say anything that was untrue. Pat's wonderful work? Yeah, the stock price has his name written ALL over it.

 

[Thunder 57]

Didn't say anything that was untrue. Pat's wonderful work? Yeah, the stock price has his name written ALL over it.

That may be, but it's not his fault investors only care about short term gains these days.

 

[SiliconFly]

.

 

[SiliconFly]

.

 

[Thunder 57]

True. Sadly, when a company goes thru a reboot, the profits/debts all take a nasty turn. But it pays off better in the future. They've invested heavily in IFS & they're gonna see the first benefits this year end. A ton of money has already been invested, they don't have to pay a truck load of money to TSMC this year. And also, their IFS profit margin comes back directly to them. Double benefit. Panther Lake & Diamond Rapids should be very high margin compared to ARL & LNL. Credit definitely goes to Pat.

I'm still in "I'll believe it when I see it" mode but there's a chance for a turnaround to be sure.

 

[OneEng2]

In any M4, Oryon 2 review it's evident Lunar is behind in efficiency and even peak 1T performance. This shouldn't be the case if Intel is to continue competing

Link?

1T performance is the primary point in mobile. In laptop, desktop, and most certainly in workstation and server, it is MT performance and PPA.

If you don't have to design much in the way of MT capabilities, the entire transistor budget can be used to enhance 1T performance. I think it is crazy to talk about ARM vs LLK in terms of only 1T.

Reading this felt like a Picard head in hands moment. Intel is using TSMC as we speak. Where's the success?! Stop looking outside for reasons this company is failing. The reasons are inside. Intel inside.

Well, when AMD moved to TSMC, they were one and a half feet in the grave (maybe more). While I agree that TSMC's process success has significantly lead to AMD's business success, it wasn't JUST AMD moving to TSMC for FAB. It was a very good Ryzen core design, spending more time on a platform (not wasting time and money constantly changing platform design), thinking strategically and more broadly (like big/medium cores that share the same uArch), and moving quickly into chiplet designs for a host of really good reasons .... and perfecting designs that work well with this kind of modular uArch.

Intel is doing lots of these things NOW ..... but AMD has been doing it longer and is currently better at it (not saying things will stay this way forever though).

Lets talk about AMD moving to TSMC a little more. Not only did this relieve AMD (which was 1/50th of Intel's size at the time) from expending huge sums of money and engineering resources trying to do BOTH fab and chip design, it let TSMC utilize economy of scale across MANY customer volumes to build a world leading FAB with the best on-boarding process for any new customer that should want to use it.

I personally think Intel trying to return to the glory days of having process supremacy (by 1 to 2 full process nodes) over all others at all times, is at an end. This model no longer makes any more business sense than a brick and mortar book store. Intel will have to spin off the foundry, and re-invent themselves with designs based on making a PROFITABLE product. No more "twice the transistor budget" over everyone else. No more "Use of monopoly power" to tilt the board their direction. No more "Soak the competition through low ball prices" using forever cash reserves.

Intel has to innovate and become much more efficient and competitive.

 

[Doug S]

In any M4, Oryon 2 review it's evident Lunar is behind in efficiency and even peak 1T performance. This shouldn't be the case if Intel is to continue competing

AMD is also behind M4, but that doesn't matter any more than Intel being behind them. AMD is Intel's main competition, not Qualcomm and definitely not Apple.

 

[cannedlake240]

In laptop, desktop, and most certainly in workstation and server, it is MT performance and PPA.

Huh? M4 consistently outperforms LNL in MT tasks when tdp constrained. Is there anything that makes Apple or x86 chips for that matter, specially optimized for either ST or MT? They're all just general purpose CPUs after all, none of them are especially large in Si area. Arm has been making progress in ST perf, but they arent a mobile only uarch designer

AMD is Intel's main competition, not Qualcomm and definitely not Apple.

Idk, Arm client designs might finally start gaining traction starting with Oryon v2 and the rumored NV soc. X elite definitely could have had a smoother release but those issues are all solvable given enough time and finances. Also AMD unlike Intel, was probably really broke until after Zen 3 launch. Client isn't their largest revenue source either

 

[cannedlake240]

Not only did this relieve AMD (which was 1/50th of Intel's size at the time) from expending huge sums of money and engineering resources trying to do BOTH fab and chip design, it let TSMC utilize economy of scale across MANY customer volumes to build a world leading FAB with the best on-boarding process for any new customer that should want to use it.

They still could've made it work had they not made the choices that led to the 10nm debacle, and if they managed to capture some of the mobile and GPU/AI businesses. I doubt it was the financial constraints that led to any of those. An argument could made about the company becoming way too big, hence complex to manage with both leading edge manufacturing and design, which in turn causes all sorts of internal issues.
Overall though Pat probably agrees with this point, which is what kick started the whole foundry business saga in the first place

 

[Thunder 57]

No one is indispensable. In every day and age, there is always a worthy successor. Whether the people in charge really want that person to lead the effort is the real question. Here's the thing: a truly worthy person will tell the Intel board things they DO NOT want to hear. Hence, they will keep appointing jackasses. But maybe this time they should go with an external jackass. Anyone other than another Intel jackass.

I think the question is, could they get someone external that's worth a crap? I can't imagine there's a bunch of CEO quality people out there who have Intel on their short list.

 

[DrMrLordX]

I think the question is, could they get someone external that's worth a crap? I can't imagine there's a bunch of CEO quality people out there who have Intel on their short list.

What exactly is a "CEO quality" person? What does Intel need from a CEO?

 

[Thunder 57]

What exactly is a "CEO quality" person? What does Intel need from a CEO?

Damned if I know. That's for the board to figure out.

 

[511]

Damned if I know. That's for the board to figure out.

If Board have that figured out they wouldn't have hired swan or fired pat midway his strategy or hired Kranzich.
Board lacks vision and even common sense
I know for sure the design team was spoiled allowing them infinite tape ins something which should never have happened there is a cultural rot at the board as well

 

[Thunder 57]

If Board have that figured out they wouldn't have hired swan or fired pat midway his strategy or hired Kranzich.
Board lacks vision and even common sense
I know for sure the design team was spoiled allowing them infinite tape ins something which should never have happened there is a cultural rot at the board as well

Swan was always going to be a "transition" CEO. Kranzich, can't say anything nice there. The board has been scrutinized lately, and IMHO rightfully so. Now they have this two headed CEO thing going on which I think is a joke. "Iceberg, right ahead!" "Hard to port, reverse full!" "No! Hard to port, increase to full!". Engine room: "???"

Obviously that's being a bit dramatic about it, but you get the idea.

OT but regarding the Titanic reference there where some different ideas as to what should've been done.

What they did was turn and reverse which seems instinctual. There are some that say they should've kept the engines forward as it would speed up the turn. Then of course they could've rammed it and despite being disabled likely would've stayed afloat. I'll end my nerd stuff here and see myself out.

 

[511]

Btw anyone with the leaked 2025-2028 Intel Roadmap ?

 

[Doug S]

Idk, Arm client designs might finally start gaining traction starting with Oryon v2 and the rumored NV soc. X elite definitely could have had a smoother release but those issues are all solvable given enough time and finances. Also AMD unlike Intel, was probably really broke until after Zen 3 launch. Client isn't their largest revenue source either

"Might"...

A lot of people here seemed to take it as a given based on Qualcomm's performance vs Intel/AMD laptop offerings that it would be a big hit. It was not. Qualcomm has taken less than 1% of the Windows PC market, and retailers have reported a high degree of product returns due to software incompatibilities. Qualcomm disputes that, but who are you going to believe, the company with a vested interest in covering it up who are two degrees removed from the consumer purchase process, or the retailers who actually handle the returns?

If Oryon v2 is faster, or Nvidia comes out with one that has a much better GPU, how does that address the reason they were returned? Would even 50% more performance than the x86 competition cause a lot of people to suddenly want to buy them, if they still have compatibility issues?

Even being first out the gate with a Windows "AI PC" before Intel/AMD released theirs did nothing for them, and that PR advantage is now gone. The dynamics of the Windows market are not going to be impacted by more or better ARM PC options. The reasons consumers are staying away are not going to change. In fact, I would argue the situation is WORSE now because developers have seen how little of an impact ARM is having on the Windows market. Some developers may have previously made tentative plans for an ARM port with their next major version, others may have taken a wait and see attitude and planned to base their decision on ARM PC sales figures.

Now that they've seen the terrible sales and return rates, it is likely lot of them have scrapped any plans for an ARM port as not being worth it given the tiny potential market. That means the long term prospects for an improvement in the reasons consumers returned them are looking pretty dim.

 

[igor_kavinski]

I can't imagine there's a bunch of CEO quality people out there who have Intel on their short list.

If they want someone like Pat to dance around the stage, I'm available. I estimate about $10 million should be enough for me to get the latest CPUs from the big players for at least a decade. I think I would be able to survive long enough to make the $10 million salary.

 

[511]

If they want someone like Pat to dance around the stage, I'm available. I estimate about $10 million should be enough for me to get the latest CPUs from the big players for at least a decade. I think I would be able to survive long enough to make the $10 million salary.

u won't be able to fix the problem permanently cause the problem is the board now not the CEO

 

[SiliconFly]

.

 

[igor_kavinski]

u won't be able to fix the problem permanently cause the problem is the board now not the CEO

Who says I'm gonna go in with the intent to fix anything?

Look, here's what I'm gonna do:

Have a throne room built.

Install at least 100 24-inch LCDs. Have feeds running on them for everything I deem important to Intel's operations. Press a button, advise someone or shout at someone or flirt with the pretty receptionist (hey, gotta know what she's up to).

Hire Peter Dinklage full time to terrorize the CPU teams (Have this done by tomorrow. You don't want me showing up here tomorrow. Trust me, everyone who knows that is up there. Yeah. UP there. *points towards the sky*).

Have them create an ARM version of Clear Linux that runs x86-64 executables in the fastest emulated way possible.

I've shared too much already. Intel can have the rest of my plans once they sign an unconditional $10 million contract with me (they can fire me whenever they want. I still walk away with $10 million).

 

[MoistOintment]

True. Same thing happened with Microsoft Windows Mobile OS. Exactly like Microsoft Windows-On-ARM.

They had a fantastic mobile OS which had better mind-share and positivity surrounding it (in the beginning). They had many phone models with fantastic hardware from different vendors (not just Nokia). i even bought one out of curiosity just to try it out. And it was pretty much perfect. Loved it. Except for Apps support which was terrible. And the rest of the story is just public knowledge. They couldn't onboard enough developers on time and it went bust.

WoA is walking the exact same path. Consumers rejected it. Developers rejected it. In spite of all the tech wizardry, emulation, etc, it just failed miserably and didn't even capture 1% of the market (and they were projecting 50% or something I think ).

WoA is a disaster. QC & MS are just beating a dead horse in the hopes no one would notice they've wasted billions on development.

The 1% figure is 1% of all Windows laptops sold in Q3 2024 were SnapDragon models.

The 50% figure is Qualcomm's goal of new PC's shipping with ARM by EOY 2029.

I highly doubt we'll see as high as 50% in 5 years - that's just overly ambitious. But the 1% figure is the first full quarter of SDE being on market.

Consumers aren't necessarily rejecting WoA because they don't know what ARM even is. They're, in large part, rejecting Qualcomm for the same reason that they're rejecting, for the most part, AMD in laptops: For the majority of average users, non-Intel is "off brand". Everyone knows Intel. Large swaths of the US population still don't recognize AMD.

But Nvidia is near a household brand at this point. And they will likely see even greater success than Qualcomm in this regard.