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.)

SiliconFly · Saturday at 4:24 PM

511 said:
I don't like him that much but he is a lot better CEOs than 3 people before him ...

True. And he's definitely 100% better than his last 2 predecessors. But as of now, Pat being "better" isn't good enough. Intel needs someone like Lisa or Jensen. Intel now needs "awesome" to survive.

gdansk said:
Do you know anything about Pat? The only reason there are still plans for discrete GPUs is because he is obsessed with Intel having their own graphics that they can eventually scale into DC. Any other CEO would have killed all those plans by now.

And it's the same Pat who fired Raja Koduri thus putting Intel Graphics division years behind competition.

511 said:
I am pretty sure you don't know much more than Intel on how to run fabs

It's common knowledge. At this point, the entire world is wondering whether Intel really knows how to run fabs considering Intel 14++++++ debacle, Intel 10+++++ debacle, Intel 4 delayed ramp and messed up hot lots, Intel 20A cancellation, 18A revised PPA and whether 18A can hit volume on time (which I seriously doubt now considering Ann Kelleher is now in the hot seat and is being setup to take all blame). Intel fabs is a disaster. If 18A fails, Intel fails. Plain and simple.

igor_kavinski said:
That's true. However, ST performance of Raptor Lake improving a bit when HT is turned off (and quite a few gamers have been doing that since Alder Lake launch) means that some resources get statically partitioned so it's not dynamic HT. Turning on HT starves the primary thread of some resources it could've had if HT hadn't been turned on.

HT is over-hyped. Instead of investing time, money & die space in HT, if they can use it to improve ST like Apple, thats the best way forward. HT on clients is more of a vestigial organ now considering the sheer no. of cores.

igor_kavinski said:
Pat's been too slow for Intel. Intel needs someone ambitious with a great sense of direction. Not some old has-been who can't figure out what Intel is good at.

You nailed it. Pat is definitely good. But not good enough. Didn't know how to manage money and has brought Intel to its knees. There are so many sharks swimming in circles around Intel trying to acquire what they can. Worst situation.

Kocicak said:
HT is not needed anymore for normal consumer workloads.

Very true.

gdansk said:
I argued that any other CEO would have cut more and earlier. It seems like a weird thing to criticize Pat for unless you are adamant he should have done so earlier. He delayed most the GPU cuts until Intel posted a billion dollar loss.

Comparisons with lesser CEOs don't mean much. Think different. Steve Jobs brought Apple back from oblivion with just his vision. So did Lisa Su. Pat pales in comparison at the moment. If 18A shines, he may have a future. Otherwise, AMD will become the next x86 monopoly.

Hitman928 said:
... but I also see some of his moves that show the "old think" at Intel that led to so many of their problems to begin with.

He's gambled way too much. Too much investments in fabs & unceremoniously pushed many of their key products into 18A without any assurances whether 18A can deliver on time. Pat might just be the straw that breaks the camel's back. If 18A fails, we'll end up with one of these (based on acquisition rumors):

Qualcomm Nova Lake
Apple Nova Lake
Samsung Nova Lake

gdansk · Saturday at 4:28 PM

SiliconFly said:
HT is over-hyped. Instead of investing time, money & die space in HT, if they can use it to improve ST like Apple, thats the best way forward. HT on clients is more of a vestigial organ now considering the sheer no. of cores.

I'm not so sure about that. If my measurements are correct Zen 5C on N3E is about the same size as Skymont on N3B but it has SMT too which helps quite a lot in certain workloads. So it seems to waste less die space than letting Intel design a CPU core.

gdansk · Saturday at 4:31 PM

AcrosTinus said:
Yes I am coping, I'm losing real money here

How so? I'm losing money on my INTC investments but I knew it was a risk. And I don't see ARL-S having much impact on their stock value. It's more the laptop parts that will matter.

jdubs03 · Saturday at 4:33 PM

SiliconFly said:
He's gambled way too much. Too much investments in fabs & unceremoniously pushed many of their key products into 18A without any assurances whether 18A can deliver on time. Pat might just be the straw that breaks the camel's back. If 18A fails, we'll end up with one of these (based on acquisition rumors):

Qualcomm Nova Lake

Apple Nova Lake

Samsung Nova Lake

I don’t think any of them will see much synergy by acquiring Intel, certainly not for CPU design. They’ve been surpassed years ago with M1. Apple put them in their place, and their place is low on the totem pole.

gdansk · Saturday at 4:35 PM

SiliconFly said:
Steve Jobs brought Apple back from oblivion with just his vision.

Just his vision? He also used his Pixar money and got billions more from Gates and Ellison in order to save Apple.

maddie · Saturday at 4:45 PM

SiliconFly said:
HT is over-hyped. Instead of investing time, money & die space in HT, if they can use it to improve ST like Apple, thats the best way forward. HT on clients is more of a vestigial organ now considering the sheer no. of cores.

Is this the new cause to rally behind? I'm seeing it expressed all over. Schizo-times. One day it's great, next it's trash.

SiliconFly · Saturday at 4:49 PM

maddie said:
Is this the new cause to rally behind? I'm seeing it expressed all over. Schizo-times. One day it's great, next it's trash.

Nope. HT in servers is awesome. HT on clients is useless. Considering the sheer no. of cores in clients these days, improving ST like Apple (and even Qualcomm) is more important. HT is over-hyped in clients.

DrMrLordX · Saturday at 5:04 PM

gdansk said:
And I don't see ARL-S having much impact on their stock value. It's more the laptop parts that will matter.

Arrow Lake-H may redeem the effort, but sadly the core won't have any immediate application in DCG where Intel could have used a crossover part. Sadly for Intel, they haven't had a successful (relative to market) new core design in DCG since Skylake. Opportunity lost.

gdansk · Saturday at 5:06 PM

SiliconFly said:
Nope. HT in servers is awesome. HT on clients is useless. Considering the sheer no. of cores in clients these days, improving ST like Apple (and even Qualcomm) is more important. HT is over-hyped in clients.

Qualcomm? Now you're being silly. Maybe you know some future product but as is all it has going for it is lower 1T power. Currently, both Intel and AMD are ahead of them in absolute 1T performance. And AMD does it on a similar core size (without throwing away SMT).

MS_AT · Saturday at 5:18 PM

Kocicak said:
You have NO idea what are you talking about. There are three types of threads:

1st thread on P core,
thread on E core
2nd thread on P core.

These threads are utilised in this order, as the thread count of the load increases.

If you have 8 thread load, all these threads will be on P cores.

Each thread above 8 will be placed on E core.

Only after you run out of unutilised cores, the second threads on P cores will be utilised.

For example for 14700K, CPU with loads with 20 and less threads will perform exactly the same whether the HT is on or off. Only the extreme 28 thread loads will utilise HT and will improve performance of these loads by about 10%.

HT is not needed anymore for normal consumer workloads.

You are talking about software abstraction that scheduler is using to manage software that doesn't manage it's own thread allocation on the simple principle that completely unused core should be better than sharing the core with someone else.
2nd P core thread is not handicapped compared to 1st one but hand-optimized software for a specific uarch will be able to utilize enough of the core that the 2nd thread will be useless. The thing is, the majority of the software, especially consumer software is not being optimized to that level, so using HT for these is net win. How big win depends on the code in question. But we can safely assume that at least Cinebench would rather benefit from HT enabled ArrowLake

moinmoin · Saturday at 5:20 PM

adroc_thurston said:
That's not really fair, AMD LLC implementation is very very very good.

How is it not fair, AMD's LLC implementation is good in parts because they started small with a 4 cores complex. The 4 cores complex of Intel's E-cores is also pretty okay. It shouldn't be news to Intel that increasing the ring size without adapting it is going to make it slow. They still did just that, repeatedly.

Intel's main problem with this launch is that they already fell on their face with Meteor Lake and skipped a desktop launch as a result of that, but falling on their face again with Arrow Lake they still proceeded with a desktop launch. All while in Lunar Lake having a design that would far much better at least on the lower end of the desktop range.

Insanity is doing the same thing over and over and expecting different results, as the saying goes.

maddie · Saturday at 5:20 PM

MS_AT said:
You are talking about software abstraction that scheduler is using to manage software that doesn't manage it's own thread allocation on the simple principle that completely unused core should be better than sharing the core with someone else.
2nd P core thread is not handicapped compared to 1st one but hand-optimized software for a specific uarch will be able to utilize enough of the core that the 2nd thread will be useless. The thing is, the majority of the software, especially consumer software is not being optimized to that level, so using HT for these is net win. How big win depends on the code in question. But we can safely assume that at least Cinebench would rather benefit from HT enabled ArrowLake

Good.

I drafted a long similar reply, but decided, why bother.

KompuKare · Saturday at 5:24 PM

maddie said:
Is this the new cause to rally behind? I'm seeing it expressed all over. Schizo-times. One day it's great, next it's trash.

I would really love to know if the extra silicon for SMT was actually taken out of Lion Cove or like AVX512 on Alder Lake it was just disabled.

Because of it was merely turned off, then the rumours that it is missing in ARL because they ran out of time to validate it might be true.

And that implies that all this SMT is bad talk all started when Intel marketing started to panic and needed some spin about why ARL would drop SMT/HT.

SiliconFly · Saturday at 6:14 PM

MS_AT said:
You are talking about software abstraction that scheduler is using to manage software that doesn't manage it's own thread allocation on the simple principle that completely unused core should be better than sharing the core with someone else.
2nd P core thread is not handicapped compared to 1st one but hand-optimized software for a specific uarch will be able to utilize enough of the core that the 2nd thread will be useless. The thing is, the majority of the software, especially consumer software is not being optimized to that level, so using HT for these is net win. How big win depends on the code in question. But we can safely assume that at least Cinebench would rather benefit from HT enabled ArrowLake

HT will definitely make a difference in benchmarks and heavily multi-threaded workloads. But general consumers don't benefit from it. HT is pretty much useless for common people. But for Cinebench MT, definitely! No questions there.

igor_kavinski · Saturday at 6:30 PM

SiliconFly said:
HT is pretty much useless for common people.

It's more useless for them coz they have no idea what HT is

We geeks and nerds just swoon at the idea of a virtual thread appearing out of nowhere like magic and thinking, "Wow! OS thinks there's twice the number of cores than they really are!". This level of elaborate trickery is second only to emulating a completely different arch and ISA on a CPU.

Let's say tomorrow Apple enhances their M5 with two ST dedicated P cores, two P cores with slightly less ST boost but ability to spawn extra threads via SMT and 4 E cores. No one's gonna mind two extra threads.

Meteor Late · Saturday at 6:30 PM

It would be nice first to know how much relative area SMT or HT has in a given core, before talking about its usefulness.

gdansk · Saturday at 6:33 PM

SiliconFly said:
HT will definitely make a difference in benchmarks and heavily multi-threaded workloads. But general consumers don't benefit from it. HT is pretty much useless for common people. But for Cinebench MT, definitely! No questions there.

But SMT allows AMD to use a single 70mm² die to compete with the 6+8 245K even in half the MT workloads. It's kinda obscene.

poke01 · Saturday at 7:07 PM

Look the only reason why Intel dropped HT is because of validation. LNC was late.
Apple currently doesn't need SMT, maybe in the future they will include it. Apple's bread and butter is iPhone, phones do not need SMT.
But Apple is building its own servers and SMT is useful there I who knows M6 or M7 might have SMT.

Intel does need SMT its core busniess is DC first then client. Maybe Unified Core will bring back HT, provided they see the worth in validading it.

reaperrr3 · Saturday at 7:11 PM

SiliconFly said:
HT will definitely make a difference in benchmarks and heavily multi-threaded workloads. But general consumers don't benefit from it. HT is pretty much useless for common people.

Nobody who held on to their 2500K for too long (like me) would ever agree to that.

adroc_thurston · Saturday at 7:44 PM

poke01 said:
Look the only reason why Intel dropped HT is because of validation. LNC was late.

Not precisely.

poke01 said:
Maybe Unified Core will bring back HT, provided they see the worth in validading it.

DC variant of LNC at Al will still feature SMT.

H433x0n · Saturday at 8:11 PM

AcrosTinus said:
Ian Cutress said it the best in his tweets, Intel was never in position to satisfy the tech reviewer crowd.

What that crowed wanted:

A CPU from Intel that is faster than the AMD 9950X

Faster than the 7800x3D and potentially significantly faster so that the upcoming 9800x3D is within rage as well.

While consuming as much energy as the 7800x3D, a chip with half its guts missing and just a fraction of the MT performance.

Yes, That was never going to happen. I think it's pretty clear that Intel needs a dedicated gaming SKU and without it, DIY desktop is basically gone forever. You cannot compete with the 7800X3D with the existing product segmentation, they're very different products. The 1T and nT performance is seemingly irrelevant DIY desktop. I would argue that it's a detriment since it means it'll consume more power due to it having more cores and higher clocks for the general compute performance.

AcrosTinus said:
You see how constrained the whole tech media space is, in any case Intel would get negative press.

Scenarios:

Intel goes for the performance again, with a higher base TDP, aggressive clocks and very high power draw. You get in all cases better performance than the 14900K

Result: It would be compared to the 7800x3D, we would get 100s of videos with a flames in the thumbnail. No one would even acknowledge the performance, all would focus on the power draw in order to negate the competitive performance while ignoring the MT performance. The fate of the 13900K and 14900K would repeat itself.

Intel goes for a imaginary skew that uses the old Intel 7 and only 8 Lion Cove cores with HT enabled and your favorite fairy dust Level 4 Cache with 100MB. Let us say it just exceeds the 7800x3D by 5%.

Result: The power draw, now of course MT performance would matter to reviewers, they will start benchmarking shader compilation.

They will remind you in every step that the platform might never get a new CPU and that AMD will last until the heat death of the galaxy. (Intel always has 2 to 3 gens and never ever has confirmed in the past)

They will start discouraging people from considering it by mentioning the 9800x3D that has not launched and plastering frame per watt graphs everywhere. There might be an additional disclaimer to tell people that even though this Intel 7 Gen3 is said to be safe every reviewer would tell people to not trust it and that it is a old 10NM node in reality that has failed already.

Intel goes for efficiency with no performance gain.

Result: mention that the platform is dead, say that AMD is better due to AMD lasting until the end of time. Compare it to the 7800x3D again, then mention the 9800x3D and tell everyone to not buy it.

DOOMSDAY:_

Go for efficiency with performance regression:

Everything mentioned above.

There was never a world where Intel would get positive media coverage but meeting the crazy standard mentioned at the start.

As they should because any imaginary world I mentioned is better than where we ended up.

I agree wholeheartedly, at this point the Intel brand is pretty toxic for the DIY desktop community and there isn't really a way to win. However, as somebody that doesn't have any built-in bias against Intel I still think ARL is a bad launch and a bad product. It's really clear why MTL-S never became a thing now.

majord · Saturday at 8:18 PM

SiliconFly said:
HT will definitely make a difference in benchmarks and heavily multi-threaded workloads. But general consumers don't benefit from it. HT is pretty much useless for common people. But for Cinebench MT, definitely! No questions there.

If this was true, all the productivity benchmarks wouldn't scale in performance from, say a 9900X to 9950X (i.e 24 to 32 threads) now wouldn't it, so with the current Core/thread count(s) on client at least it absolutely is not useless.

gdansk · Saturday at 8:20 PM

1851 has no future (or at least Intel is unwilling to commit to saying it has a future).
Arrow Lake S can't game. It trades blows with the competition in MT. It uses more power than the competition whenever it actually does something other than sit idle. Somehow it sucks at web browsing. Reviewers didn't publish results with the current build of Windows because it is broken. Many mention numerous crashes.

And yet some people want to act like the poor reception is the tech media being unreasonable to Intel. No, Intel shipped this. They're being charitable jumping through Intel's hoops to find a bright side. In its present state it's a charity case. Why spend more on getting this instead of the 14900K or 9950X?

H433x0n · Saturday at 8:43 PM

gdansk said:
1851 has no future (or at least Intel is unwilling to commit to saying it has a future).
Arrow Lake S can't game. It trades blows with the competition in MT. It uses more power than the competition whenever it actually does something other than sit idle. Somehow it sucks at web browsing. Reviewers didn't publish results with the current build of Windows because it is broken. Many mention numerous crashes.

And yet some people want to act like the poor reception is the tech media being unreasonable to Intel. No, Intel shipped this. They're being charitable jumping through Intel's hoops to find a bright side. In its present state it's a charity case. Why spend more on getting this instead of the 14900K or 9950X?

Do we actually know it’s less efficient? I haven’t seen any mainstream testing data except for nT performance at PL2.

The only graphs I’ve seen of a perf/watt curve shows ARL doing better in power limited scenarios in cinebench R23 & R24.

gdansk · Saturday at 8:48 PM

H433x0n said:
The only graphs I’ve seen of a perf/watt curve shows ARL doing better in power limited scenarios in cinebench R23 & R24.

The curves we have for Cinebench R23 show it is efficient only at the TDPs they didn't ship it at.
And for SPECint it looks worse all along.

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

Senior member

Attachments

Golden Member

Diamond Member

Diamond Member

Senior member

Diamond Member

Diamond Member

Golden Member

Lifer

Diamond Member

Senior member

Diamond Member

Diamond Member

Golden Member

Golden Member

Lifer

Member

Diamond Member

Platinum Member

Junior Member

Diamond Member

Golden Member

Senior member

Diamond Member

Golden Member

Diamond Member