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

[Tigerick]

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	8 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 (20A)	Arrow Lake (N3B)	Arrow Lake Refresh (N3B)	Lunar Lake	Panther Lake
Platform	Mobile H/U Only	Desktop Only	Desktop & Mobile H&HX	Desktop Only	Mobile U Only	Mobile H
Process Node	Intel 4	Intel 20A	TSMC N3B	TSMC N3B	TSMC N3B	Intel 18A
Date	Q4 2023	Q1 2025 ?	Desktop-Q4-2024 H&HX-Q1-2025	Q4 2025 ?	Q4 2024	Q1 2026 ?
Full Die	6P + 8P	6P + 8E ?	8P + 16E	8P + 32E	4P + 4E	4P + 8E
LLC	24 MB	24 MB ?	36 MB ?	?	8 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]

But you yourself suggested measuring MTL in an idle, non-performance sensitive scenario. Why is performance important in scenarios when you're barely touching a keyboard (the only ones that LP-E cores are actually doing anything)?

Funny how the words are easily misinterpreted. Now please try to understand it clearly so that we're on the same page. LPE cores in MTL are more like E cores in Apple M-series processors. They perform the same function.

And my recommendation is to test MTL production laptops with the BIOS patch (with pcode update) on real world use cases if we want to gauge it's real power efficiency. Not idle like what you're implying. The difference is huge. The answers are already here in this forum as this discussion has been going on for quite a while. Please feel free to go thru it.

And regarding Windows' 11 stability: I guess my ROG Ally didn't get the memo about its stability when it crashed into BSOD while playing Genshin Impact, failing to boot afterwards and forcing me to do cloud recovery that resulted in the Windows 11 being reinstalled without a freaking Microsoft Store.

And I've had far worse experiences with my mac. Best of luck.

 

[Abwx]

I tend to believe GoldenPig & UltraBookReview know how to review products and take accurate measurements. However, it’s possible they both screwed up. I don’t think it’s the most likely scenario but who knows.

We’ll see when more people get their hands on production laptops with the updated Bios.

Methink that they noticed that focusing on perf/Watt was counterproductive, not only MTL is not more efficient than the competition but it appeared also as slower.

So they came with this patch that worsen perf/watt but allow for bigger numbers in the benchs, they didnt say what are the changes in this bios so people are led to think that there s a perf/watt improvement while it s the opposite, that s how deceptive marketing work...

 

[SiliconFly]

Methink that they noticed that focusing on perf/Watt was counterproductive, not only MTL is not more efficient than the competition but it appeared also as slower.

So they came with this patch that worsen perf/watt but allow for bigger numbers in the benchs, they didnt say what are the changes in this bios so people are led to think that there s a perf/watt improvement while it s the opposite, that s how deceptive marketing work...

I don't think fixing BIOS bugs is deceptive. Actually, it's their responsibility. I know MTL beating some obscure AMD chips like 7840 might make some unhappy, but thats what it is. MTL is already far more efficient than the previous gen it replaces and is already inline or exceeding competition and is poised to outsell it by a very very wide margin.

 

[igor_kavinski]

And regarding Windows' 11 stability: I guess my ROG Ally didn't get the memo about its stability when it crashed into BSOD while playing Genshin Impact, failing to boot afterwards and forcing me to do cloud recovery that resulted in the Windows 11 being reinstalled without a freaking Microsoft Store.

TBH, that sounds more like a ASUS hardware issue.

 

[igor_kavinski]

I know MTL beating some obscure AMD chips like 7840 might make some unhappy

Obscure??? LMAO!

 

[cebri1]

Core Ultra 5 125H DTT On vs DTT Off

In 3 games DTT off runs faster, 1 game DTT on faster, 1 game equal. GPU clock speeds reporting is wrong or not? Seems low, not sure. CPU clock is wrong for sure.

On MTL devices with LP 7500 the memory can be downclocked to 5600 by the looks of it. SAGV seems to be off by default which is faster but also not as efficient. If it's off in battery mode this might be not so great.

Am I reading this right? is it peaking at 25-30W on most games? I would be expecting a higher power draw specially on games like Howartz Legacy

 

[Hulk]

You also get to access the OS that does not BSOD on you (and spies just as well).

I take umbrage at this assessment. There is no better spyware than Windows.

 

[mikk]

Am I reading this right? is it peaking at 25-30W on most games? I would be expecting a higher power draw specially on games like Howartz Legacy

Samsung Galaxy Book4 is limited to this.

 

[Abwx]

I don't think fixing BIOS bugs is deceptive. Actually, it's their responsibility. I know MTL beating some obscure AMD chips like 7840 might make some unhappy, but thats what it is. MTL is already far more efficient than the previous gen it replaces and is already inline or exceeding competition and is poised to outsell it by a very very wide margin.

That could be written by a deceptive marketing department...

Acually MTL beat nothing despite a 6 + 8 having a theorical 25% equivalent core count advantage over a 7840, that s telling that Intel 4 is not even up to TSMC s N5 and closer to N6.

Prove is its TSMC s N6 fabbed GPU that has comparable perf/Watt to the 780M thanks to 25% higher SP count that compensate for the process difference.

 

[cebri1]

Samsung Galaxy Book4 is limited to this.

Then pretty acceptable results at those power levels. The 8Xe Core SKUs should provide a nice bump.

 

[H433x0n]

That could be written by a deceptive marketing department...

Acually MTL beat nothing despite a 6 + 8 having a theorical 25% equivalent core count advantage over a 7840, that s telling that Intel 4 is not even up to TSMC s N5 and closer to N6.

Prove is its TSMC s N6 fabbed GPU that has comparable perf/Watt to the 780M thanks to 25% higher SP count that compensate for the process difference.

The iGPU is on N5P. I’ll make sure to revisit this once more outlets get a hold of laptops with production pcode. At that point I’ll predict that you’ll move onto saying it’s not a fair comparison since it’s 6P+8E vs 8P (despite the MTL cpu tile being same size as a Zen 4 CCD).

 

[SiliconFly]

That could be written by a deceptive marketing department...

Looks like your post is written by some cheap and deceptive marketing department. Reeks of desperation about a failed product like 7840. (Recommend not to engage in personal attacks. As recommended by the mods, attack the post not the person who posts.)

...telling that Intel 4 is not even up to TSMC s N5 and closer to N6...

Shows that you don't understood fabs at all. And I'll repeat what I mentioned earlier, AnandTech, Toms Hardware, SemiWiki & WikiChip have all rated Intel 4 to be better than TSMC N4 and closer to N3. You can believe what ever you want, but facts are facts.

 

[FlameTail]

Intel 4 to be better than TSMC N4 and closer to N3. You can believe what ever you want, but facts are facts.

In what aspect? Performance? Power? Area?

 

[Abwx]

Shows that you don't understood fabs at all. And I'll repeat what I mentioned earlier, AnandTech, Toms Hardware, SemiWiki & WikiChip have all rated Intel 4 to be better than TSMC N4 and closer to N3. You can believe what ever you want, but facts are facts.

Then why is Intel using TSMC s N5 to fab the GPU part if Intel 4 was better than TSMC s N4..?..

 

[SiliconFly]

In what aspect? Performance? Power? Area?

It's a very old topic thats not under discussion anymore. You can find more info by searching for the Intel 4 article in wikichip & semiwiki that have very detailed analysis of the node. Too long to read, but they very clearly specify what you want to know.

A snippet from semiwiki:
TSMC N5 has a 51nm CPP and 34nm M2P with a 9.00 track high-performance cell that yields a 306nm CH and a 15,606nm2 CPP x CH. We believe TSMC N3 has a 45nm CPP and 28nm M2P, and for a 9.00 track high-performance cell that yields a CH of 252nm and a CPP x CH of 11,340nm2. For Intel 4 the CPP is 50nm and M2P is 45nm (disclosed in the briefing although not in the paper), this yields a tracks height of only 5.33 for the quoted 240nm CH and a CPP x CH of 12,000nm2. These values are consistent with a 4 designation since it slots between N5 and N3 for the leading foundry company TSMC, although it is closer to TSMC N3 than TSMC N5. We also believe Intel 4 will have performance slightly better than TSMC N3. I didn’t include Samsung in Figure 4 but based on my current estimates Intel 4 is denser than Samsung GAE3. Samsung may have a small performance advantage over Intel 4 and TSMC N3, but Intel 3 should surpass both Samsung GAE3 and TSMC N3 for performance next year.

Both wikichip & semiwiki articles are excellent. Please take some time to read it. You may find them interesting.

Then why is Intel using TSMC s N5 to fab the GPU part if Intel 4 was better than TSMC s N4..?..

Oh god! This has been discussed at length already in this very forum. I'll repeat what has already been said by many members already just for you. Intel 4 doesn't have volume and I don't think they're gonna increase capacity as it's an intermediate node that'll be phased out soon. It's a purpose built node for MTL CPU tile alone.

 

[Abwx]

It's a very old topic thats not under discussion anymore. You can find more info by searching for the Intel 4 article in wikichip & semiwiki that have very detailed analysis of the node. Too long to read, but they very clearly specify what you want to know.

A snippet from semiwiki:
TSMC N5 has a 51nm CPP and 34nm M2P with a 9.00 track high-performance cell that yields a 306nm CH and a 15,606nm2 CPP x CH. We believe TSMC N3 has a 45nm CPP and 28nm M2P, and for a 9.00 track high-performance cell that yields a CH of 252nm and a CPP x CH of 11,340nm2. For Intel 4 the CPP is 50nm and M2P is 45nm (disclosed in the briefing although not in the paper), this yields a tracks height of only 5.33 for the quoted 240nm CH and a CPP x CH of 12,000nm2. These values are consistent with a 4 designation since it slots between N5 and N3 for the leading foundry company TSMC, although it is closer to TSMC N3 than TSMC N5. We also believe Intel 4 will have performance slightly better than TSMC N3. I didn’t include Samsung in Figure 4 but based on my current estimates Intel 4 is denser than Samsung GAE3. Samsung may have a small performance advantage over Intel 4 and TSMC N3, but Intel 3 should surpass both Samsung GAE3 and TSMC N3 for performance next year.

Both wikichip & semiwiki articles are excellent. Please take some time to read it. You may find them interesting.


Oh god! This has been discussed at length already in this very forum. I'll repeat what has already been said by many members already just for you. Intel 4 doesn't have volume and I don't think they're gonna increase capacity as it's an intermediate node that'll be phased out soon. It's a purpose built node for MTL CPU tile alone.

Semiwiki article is from june 2022 and is about expectations, not actual silicon.

If Intel 4 was to be better than TSMC s N5 as said in the article then a 6 + 8 MTL would eat a 7840 for breakfeast in anything multithreaded perf/watt wise, obviously the expectations were above the final result.

 

[lightisgood]

If Intel 4 was to be better than TSMC s N5 as said in the article then a 6 + 8 MTL would eat a 7840 for breakfeast in anything multithreaded perf/watt wise, obviously the expectations were above the final result.

This is not a control experiment...

 

[FlameTail]

TSMC CEO said N3P is competitive with Intel 18A.

Intel CEO said 18A is competitive with TSMC N2.

 

[Tarkin77]

TSMC CEO said N3P is competitive with Intel 18A.

Intel CEO said 18A is competitive with TSMC N2.

I trust TSMC, Intel is usually to "optimistic" about their products (history of overpromissing and underdelivering)

 

[TESKATLIPOKA]

The iGPU is on N5P. I’ll make sure to revisit this once more outlets get a hold of laptops with production pcode. At that point I’ll predict that you’ll move onto saying it’s not a fair comparison since it’s 6P+8E vs 8P (despite the MTL cpu tile being same size as a Zen 4 CCD).

So what If MTL CPU tile has the same size as Zen4 CCD? What does It have to do with MTL vs PHX monolith?
Even If @Abwx would say that It's unfair, so what? He is not lying, but It's also not Intel's fault but AMDs.
Didn't the other side also say that It's unfair because AMD uses a better process or P-core is just too inefficient? But that is not AMD's fault but Intel's.
Unfair or fair, we will still compare based on what's already released.

 

[TESKATLIPOKA]

Looks like your post is written by some cheap and deceptive marketing department. Reeks of desperation about a failed product like 7840. (Recommend not to engage in personal attacks. As recommended by the mods, attack the post not the person who posts.)

7840 is a failed product? Can you stop posting lies? Thank you.

 

[TESKATLIPOKA]

If Intel 4 was to be better than TSMC s N5 as said in the article then a 6 + 8 MTL would eat a 7840 for breakfeast in anything multithreaded perf/watt wise, obviously the expectations were above the final result.

Intel 4 could be competitive against TSMC N5, yet that doesn't mean MTL would automatically defeat 7840 by a large margin.
You should realize that we are comparing different cores, so that's already the biggest variable there is.

 

[poke01]

Node does matter but having a great architecture also matter even more.

Would M1 have been great with a Cortex X1 chip on TSMC 5nm?

Nope. In the end it’s the architecture and the software combined that make the experience great. These next few years are not about who’s on the best node but who has the best designs. x86 is not going anywhere but also its best to keep in mind the best innovation happens in a competitive environment and the CPU industry is very healthy now.

 

[Abwx]

Intel 4 could be competitive against TSMC N5, yet that doesn't mean MTL would automatically defeat 7840 by a large margin.
You should realize that we are comparing different cores, so that's already the biggest variable there is.

Even if the designs are different MTL has 6P, and the 8E should more than compensate for the 7840 using 8P, MTL is supposed to be on par with a theorical 10P CPU.

At the very least it could be clocked such that it would have vastly better perf/Watt at same perfs assuming an equal process is used, and since this didnt happen it means that the culprit is the process, Intel s P core could be less efficient than a Zen 4 core but not to this extent.

 

[rtxtwt]

We’ll see when more people get their hands on production laptops with the updated Bios.

So they came with this patch that worsen perf/watt but allow for bigger numbers in the benchs, they didnt say what are the changes in this bios so people are led to think that there s a perf/watt improvement while it s the opposite, that s how deceptive marketing work...

What I just heard is the new BIOS is to lower the voltage and boost higher, considering ASUS has bad track record that set voltage too high and burn Alderlake/Zen4, I guess it's just only ASUS has this problem, that being said it's not Intel's fault, and those reviewers did nothing wrong.