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

Tigerick · Aug 22, 2022

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.

Comparison of upcoming Intel's U-series CPU: Core Ultra 100U, Lunar Lake and Panther Lake

Model	Code-Name	Date	TDP	Node	Tiles	Main Tile	CPU	LP E-Core	LLC	GPU	Xe-cores
Core Ultra 100U	Meteor Lake	Q4 2023	15 - 57 W	Intel 4 + N5 + N6	4	tCPU	2P + 8E	2	12 MB	Intel Graphics	4
?	Lunar Lake	Q4 2024	17 - 30 W	N3B + N6	2	CPU + GPU & IMC	4P + 4E	0	12 MB	Arc	8
?	Panther Lake	Q1 2026 ?	?	Intel 18A + N3E	3	CPU + MC	4P + 8E	4	?	Arc	12

Comparison of die size of Each Tile of Meteor Lake, Arrow Lake, Lunar Lake and Panther Lake

	Meteor Lake	Arrow Lake (N3B)	Lunar Lake	Panther Lake
Platform	Mobile H/U Only	Desktop & Mobile H&HX	Mobile U Only	Mobile H
Process Node	Intel 4	TSMC N3B	TSMC N3B	Intel 18A
Date	Q4 2023	Desktop-Q4-2024 H&HX-Q1-2025	Q4 2024	Q1 2026 ?
Full Die	6P + 8P	8P + 16E	4P + 4E	4P + 8E
LLC	24 MB	36 MB ?	12 MB	?
tCPU	66.48
tGPU	44.45
SoC	96.77
IOE	44.45
Total	252.15

Intel Core Ultra 100 - Meteor Lake

As mentioned by Tomshardware, TSMC will manufacture the I/O, SoC, and GPU tiles. That means Intel will manufacture only the CPU and Foveros tiles. (Notably, Intel calls the I/O tile an 'I/O Expander,' hence the IOE moniker.)

moinmoin · Monday at 6:13 AM

DavidC1 said:
Other guys say the problems aren't limited to Intel. AMD is seriously behind too.

DavidC1 said:
The future of x86 is still in doubt.

Both Intel and AMD are essentially hiding behind the remainder of monopolies x86 still enjoys in many markets, Intel in laptop and enterprise, AMD in server markets. But that's not going to work well for much longer.

cebri1 said:
But yeah, it's 30 month since he took over... LOL

It's been 44 months now. For comparison Swan was CEO for 31 months, Krzanich for 61, Otellini for 96.

majord · Monday at 6:23 AM

Kocicak said:
You wrote a theoretical statement and what you disagree with what I wrote, without even a single reason why.

Can you find even ONE heavy MT consumer application? I am not talking about professional video production software, scientific calculation/simulation apps etc.

What makes you think video editing, photo editing / effects filters and rendering software are just for professional use? Is my 10 y/o a professional?

handbrake, Blender, After effects, are all very popular , they're not professional only app's by a long shot , they all benefit from MT quite heavily in some cases.

The rest of this post screams "performance doesn't matter" , in which case, why are you even interested in high performance computing as a topic at all.? The sort of questions you're asking there could have applied since the dawn of personal computing.

cebri1 · Monday at 6:26 AM

moinmoin said:
Both Intel and AMD are essentially hiding behind the remainder of monopolies x86 still enjoys in many markets, Intel in laptop and enterprise, AMD in server markets. But that's not going to work well for much longer.

It's been 44 months now. For comparison Swan was CEO for 31 months, Krzanich for 61, Otellini for 96.

You didn’t understand my post.

poke01 · Monday at 6:28 AM

good review of 245k

igor_kavinski · Monday at 6:44 AM

moinmoin said:
It's been 44 months now. For comparison Swan was CEO for 31 months, Krzanich for 61, Otellini for 96.

Clock's ticking. Intel board has to choose between Intel or Pat.

511 · Monday at 7:56 AM

moinmoin said:
It's been 44 months now. For comparison Swan was CEO for 31 months, Krzanich for 61, Otellini for 96.

Swan was ousted for pursuing stronger relationship with TSMC if anything Kranzich should have been fired sooner and pat was bought in sooner Intel would not have this much issue

igor_kavinski · Monday at 8:31 AM

igor_kavinski said:
Clock's ticking. Intel board has to choose between Intel or Pat.

Relevant to that and old news but a board member already quit thanks to Pat's unwillingness to take the needed action(s): https://www.reuters.com/technology/...nces-over-chipmakers-revival-plan-2024-08-27/

Little by little, everyone sensible is going to jump ship and all that will be left behind will be fools nodding their heads at each other's stupid hyperbole. And trust me, these idiots won't be the losers when Intel suffers another devastating quarter. They will still be well off due to their millions in their bank accounts. If and when the US Govt is forced to bail out Intel, they should ensure to fire the entire Intel board and remake it (if possible).

moinmoin · Monday at 8:33 AM

511 said:
Swan was ousted for pursuing stronger relationship with TSMC if anything Kranzich should have been fired sooner and pat was bought in sooner Intel would not have this much issue

I remember Swan didn't want to become CEO at first, and it took Intel a lot of time and money to convince Pat to come back to Intel. In general it seemed Intel had a hard time to find suitable replacements which may also be why Krzanich kept his post that long even after the affair for which he had to leave was known for a long time already.

In any case Pat is at Intel long enough already that even with all the previous decisions forcing his hands he should have enough room by now to make better strategical choices like spending already ordered TSMC N3B silicon on something more worthwhile than Core Ultra 200S series dies.

igor_kavinski · Monday at 8:34 AM

511 said:
Swan was ousted for pursuing stronger relationship with TSMC

Is that really what happened? Any URL?

511 said:
pat was bought in sooner Intel would not have this much issue

You are right. Not this much issue. Just bigger issues!

igor_kavinski · Monday at 8:35 AM

moinmoin said:
In any case Pat is at Intel long enough already that even with all the previous decisions forcing his hands he should have enough room by now to make better strategical choices like spending already ordered TSMC N3B silicon on something more worthwhile than Core Ultra 200S series dies.

What would you do?

I would issue orders to glue together two or more Lunar Lake dies

moinmoin · Monday at 8:45 AM

igor_kavinski said:
What would you do?

I would issue orders to glue together two or more Lunar Lake dies

Well, when it became clear that Meteor Lake wouldn't fly on desktop, the slight upgrade over it that is Arrow Lake surely shouldn't have gotten the green light for a desktop release either. More Lunar Lake would have been more sensible. Work on finally improving/replacing the ring with something more scalable while more efficient than the mesh would have been more sensible. Intel managed to split up its monoliths without really adding any flexibility to its designs which is in stark contrast to all its competition.

Kocicak · Monday at 8:49 AM

majord said:
handbrake, Blender, After effects, are all very popular , they're not professional only app's by a long shot , they all benefit from MT quite heavily in some cases.

Benefit heavily, you say? OK, demonstrate.

BTW if somebody wants to regularly use some of these apps, they may pick up 20 or 24 core Intel parts, so please use some of those for the demonstration.

Meteor Late · Monday at 8:55 AM

I mean, there are plenty of tests out there that are not really benchmarks in the sense that only a score is shown, but tests where you can see time needed to render a video in software like Handbrake.

OneEng2 · Monday at 9:20 AM

Kocicak said:
You wrote a theoretical statement and what you disagree with what I wrote, without even a single reason why.

Can you find even ONE heavy MT consumer application? I am not talking about professional video production software, scientific calculation/simulation apps etc.

And if you can find such an app, can you run it with and without HT, compare the results and in a realistic use scenario explain how does this difference matter to the consumer?

We are talking about current Intel consumer CPUs, which have at least 12 physical cores even in lower price category.

So cut the theoretical fluff, and talk facts:

CPU used - current Intel with at least 12 cores:
Application used:
HT benefit:
Typical workload size:
Average time spared by using HT:
Use scenario - how many times a month is such a workload processed:
Benefit to the consumer - time spared per month:
How does this benefit the customer?
Can he even notice?
How does this saved time compare to the time wasted by different unproductive activities during the month?

Wow. So many things wrong with this that it is hard to start.

I won't bother reposting what @majord wrote but to say I produce videos for my family many times a year as well as doing videos for other purposes from time to time. Also, added to this is the fact that at any given time, your desktop computer is running tons of tasks in the background .... in parallel.

Finally, designing a single core processor design that works well in both consumer and DC is a great business strategy as it is efficient from a business standpoint.

If performance doesn't matter to you, just go get yourself a used Bulldozer or Core 2 off of eBay. Making the case that performance doesn't matter in a new processor architecture is just silly.

majord said:
What makes you think video editing, photo editing / effects filters and rendering software are just for professional use? Is my 10 y/o a professional?

handbrake, Blender, After effects, are all very popular , they're not professional only app's by a long shot , they all benefit from MT quite heavily in some cases.

The rest of this post screams "performance doesn't matter" , in which case, why are you even interested in high performance computing as a topic at all.? The sort of questions you're asking there could have applied since the dawn of personal computing.

Thanks for saving me some typing .

511 · Monday at 9:29 AM

moinmoin said:
I remember Swan didn't want to become CEO at first, and it took Intel a lot of time and money to convince Pat to come back to Intel. In general it seemed Intel had a hard time to find suitable replacements which may also be why Krzanich kept his post that long even after the affair for which he had to leave was known for a long time already.

In any case Pat is at Intel long enough already that even with all the previous decisions forcing his hands he should have enough room by now to make better strategical choices like spending already ordered TSMC N3B silicon on something more worthwhile than Core Ultra 200S series dies.

It took roughly 4-5 years for Swan decisions to show in Meteor Lake/Arrow Lake/Granite Rapids Pat's will be similar cadence.

The most important will be 18A and the products on it, iirc Lunar Lake was after pat joined same with Sierra Forest.You are trying to portray him taking every wrong decision.He has taken better decisions than predecessor but some of his decisions are 🤮

Josh128 · Monday at 9:31 AM

OneEng2 said:
Wow. So many things wrong with this that it is hard to start.

I won't bother reposting what @majord wrote but to say I produce videos for my family many times a year as well as doing videos for other purposes from time to time. Also, added to this is the fact that at any given time, your desktop computer is running tons of tasks in the background .... in parallel.

Finally, designing a single core processor design that works well in both consumer and DC is a great business strategy as it is efficient from a business standpoint.

If performance doesn't matter to you, just go get yourself a used Bulldozer or Core 2 off of eBay. Making the case that performance doesn't matter in a new processor architecture is just silly.

Thanks for saving me some typing .

Handbrake / video encoding, audio encoding, OBS, compiling are some great examples of MT workloads.

moinmoin · Monday at 9:34 AM

511 said:
It took roughly 4-5 years for Swan decisions to show in Meteor Lake/Arrow Lake/Granite Rapids Pat's will be similar cadence the foremost will be 18A and the products on it btw iirc Lunar Lake was after pat joined same with Sierra Forest you are trying to force every wrong decision on him he has taken better decisions than predecessor but his some decisions are 🤮

Please use punctuation.

igor_kavinski · Monday at 9:35 AM

Josh128 said:
Handbrake / video encoding, audio encoding, OBS, compiling are some great examples of MT workloads.

Browsing with multiple tabs open is another. Try doing that on a PC with low amount of cores or shutting off HT on a 4 core CPU and only then one understands what HT does. HT mostly benefits threaded workloads, except for workloads where cache size is so important that turning off HT enables the remaining cores to finish their work faster.

Hulk · Monday at 9:59 AM

Kocicak said:
You wrote a theoretical statement and what you disagree with what I wrote, without even a single reason why.

Can you find even ONE heavy MT consumer application? I am not talking about professional video production software, scientific calculation/simulation apps etc.

And if you can find such an app, can you run it with and without HT, compare the results and in a realistic use scenario explain how does this difference matter to the consumer?

We are talking about current Intel consumer CPUs, which have at least 12 physical cores even in lower price category.

So cut the theoretical fluff, and talk facts:

CPU used - current Intel with at least 12 cores:
Application used:
HT benefit:
Typical workload size:
Average time spared by using HT:
Use scenario - how many times a month is such a workload processed:
Benefit to the consumer - time spared per month:
How does this benefit the customer?
Can he even notice?
How does this saved time compare to the time wasted by different unproductive activities during the month?

I think Kocicak is being taken out of context here.
The point being made is that modern Intel CPU's with 12 or more cores don't really benefit from HT except maybe for a few niche cases.
This has certainly been my experience with my 14900K. The 16 physical E cores are much more useful than the 8 logical cores, which only serve to heat the core beyond what my cooler can handle.

Also ARL is proof. No HT and very good MT performance. Ironically it is more the applications that don't employ all of the cores are the ones that don't perform as well as we'd hoped.

Markfw · Monday at 10:48 AM

Hulk said:
I think Kocicak is being taken out of context here.
The point being made is that modern Intel CPU's with 12 or more cores don't really benefit from HT except maybe for a few niche cases.
This has certainly been my experience with my 14900K. The 16 physical E cores are much more useful than the 8 logical cores, which only serve to heat the core beyond what my cooler can handle.

Also ARL is proof. No HT and very good MT performance. Ironically it is more the applications that don't employ all of the cores are the ones that don't perform as well as we'd hoped.

Intel never used HT to near the extent AMD, so losing it does not make that much difference performance wise, and possibly saves a lot of security concerns and problems and save some transistors for less die space. AMD does not have near the security concerns regarding HT, but loses a lot of performance. That almost makes that argument of no concern for Intel.

MarkizSchnitzel · Monday at 10:49 AM

I feel like you guys have very different definitions of what an "average consumer" looks like which is causing some friction.

If someone forced my hand, I would say the height for an average consumer is paying for bills via online banking, maybe.
Everything has moved to phones.

But, when someone buys a desktop computer, that alone makes it not average. Everyone I know personally who has a desktop is using it for gaming.
My social circle is not representative though.

Markfw · Monday at 10:52 AM

MarkizSchnitzel said:
I feel like you guys have very different definitions of what an "average consumer" looks like which is causing some friction.

If someone forced my hand, I would say the height for an average consumer is paying for bills via online banking, maybe.
Everything has moved to phones.

But, when someone buys a desktop computer, that alone makes it not average. Everyone I know personally who has a desktop is using it for gaming.
My social circle is not representative though.

See that also applies to age group. I live and breath by desktop, for many things, and can barely see whats going on, on my cell phone. My son lives and breathes for his cell phone, but uses desktop for work in CAD.

igor_kavinski · Monday at 11:12 AM

Most people with desktops and even laptops in their homes are typically "serious" users. They push their CPUs pretty hard. And these users keep the PC industry going because them pushing their CPUs makes them desire for faster and faster CPUs. Sure, Intel can get rid of HT if it allows them to surpass Zen 5 by at least 5% in EVERY workload. That's not the case with Arrow Lake so disabling or axing HT in the design phase was a tragically premature decision, possibly signed off on by the wise and great Pat, formerly known as The Larrabee Disaster Artist.

Meteor Late · Monday at 11:30 AM

Hulk said:
I think Kocicak is being taken out of context here.
The point being made is that modern Intel CPU's with 12 or more cores don't really benefit from HT except maybe for a few niche cases.
This has certainly been my experience with my 14900K. The 16 physical E cores are much more useful than the 8 logical cores, which only serve to heat the core beyond what my cooler can handle.

Also ARL is proof. No HT and very good MT performance. Ironically it is more the applications that don't employ all of the cores are the ones that don't perform as well as we'd hoped.

But all of that depends on how much cost HT or SMT adds? we know an E core is like 1/3 of a P core in space, how much is the area cost of HT/SMT? and the power efficiency and performance too of course. Only then it makes sense to compare. I've read it's about 5% of the P core area, though of course it may depend on which exact core.

On the surface, HT/SMT seems more efficient per area than E cores, that's in the case one has to choose between one or the other of course, you can do both which should already be the case with Lion Cove. But then HT also limits you very quickly, in the fact that if you have 8 cores, and add HT, then you cannot get more than the 25-30% extra you get, unlike E cores that you can keep adding more of them and gain a lot of performance.

Hulk · Monday at 11:35 AM

igor_kavinski said:
Most people with desktops and even laptops in their homes are typically "serious" users. They push their CPUs pretty hard. And these users keep the PC industry going because them pushing their CPUs makes them desire for faster and faster CPUs. Sure, Intel can get rid of HT if it allows them to surpass Zen 5 by at least 5% in EVERY workload. That's not the case with Arrow Lake so disabling or axing HT in the design phase was a tragically premature decision, possibly signed off on by the wise and great Pat, formerly known as The Larrabee Disaster Artist.

I don't understand how HT would have made ARL better?
As I've written before my HT enabled CPU was a 3.06GHz Northwood P4. HT was amazing. From one thread to two. Logical or not, that second thread kept the train moving.

E cores, espeically when you have 16 of them, have rendered logical cores moot.

Markfw said:
Intel never used HT to near the extent AMD, so losing it does not make that much difference performance wise, and possibly saves a lot of security concerns and problems and save some transistors for less die space. AMD does not have near the security concerns regarding HT, but loses a lot of performance. That almost makes that argument of no concern for Intel.

True, but it hasn't been a blow out. For example, from my CB R23 MT data for HT/SMT below while AMD has had better implimentations, generation-to-generation Intel has only been a few percentage points behind.

Zen 5 +38%

Raptor Cove +35%
Zen 4 +34%

Golden Cove +30%
Cypress Cove +26%
Zen 3 +31%
Zen 2 +31%

Zen + 25%
Zen +25%
Skylake +26%

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

Senior member

Attachments

Diamond Member

Senior member

Senior member

Platinum Member

Lifer

Senior member

Lifer

Diamond Member

Lifer

Lifer

Diamond Member

Golden Member

Member

Member

Senior member

Senior member

Diamond Member

Lifer

Diamond Member

Moderator Emeritus, Elite Member

Senior member

Moderator Emeritus, Elite Member

Lifer

Member

Diamond Member