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

 

[mikk]

In Specint they used icc putting others at a disadvantage. Also the 50% power reduction is in one benchmark, most real workloads only got about 30% power savings. Seems like Teams is the best case with 41% but it's still 20% behind X elite. 4K AV1 streaming they're claiming 33%, which based on MTL's 17W number translates to ~11W. That's still 3X of what M3 based MB Air draws and double the amount of power X Elite elite uses in this workload

Do you refer to soc power or total power usage? And from where do you get these numbers? In their AV1 4K demo MTL draw something like 12 watts total. Soc power is a fraction of this.

 

[DavidC1]

Do you refer to soc power or total power usage? And from where do you get these numbers? In their AV1 4K demo MTL draw something like 12 watts total. Soc power is a fraction of this.

Exactly, and SoC power reduction propagates into rest of the system because it's a controller for the rest.

Youtube 4K AV1 demo shows the power consumption is like 5W for Lunarlake at the system level, not package level.

 

[cannedlake240]

Do you refer to soc power or total power usage? And from where do you get these numbers? In their AV1 4K demo MTL draw something like 12 watts total. Soc power is a fraction of this.

The power number is from notebookcheck X elite efficiency article where they used an external display. In intels demo both MTL LNL laptop screens are also off.

 

[mikk]

The power number is from notebookcheck X elite efficiency article where they used an external display. In intels demo both MTL LNL laptop screens are also off.

Do you have a link? Something is off or missing.

 

[DavidC1]

Do you have a link? Something is off or missing.

Qualcomm Snapdragon X Elite Analysis - More efficient than AMD & Intel, but Apple stays ahead

Notebookcheck analysis of the new Qualcomm Snapdragon X Elite ARM processor X1E-78-100 and X1E-80-100.

www.notebookcheck.net

He's not necessarily wrong. But between two Elite systems there's quite a bit of power difference, which could bring Meteorlake numbers down to what Intel shown.

 

[mikk]

For Netflix streaming Lunar Lake can work 44% longer (26 hours) than MTL with 18 hours says Dell. The Qualcomm device can do 27 hours by the way.

For example, Dell is just now announcing a new version of its flagship $1,400 XPS 13 laptop with Lunar Lake that’s basically identical to the current model in every other way: same chassis, same dimensions, same screen, same 55 watt-hour battery pack. And yet, Dell says it now gets up to 26 hours of 1080p Netflix streaming at 150 nits of brightness, up from
18 hours previously.

That’s a 44 percent improvement, which is a giant leap for Intel, if true. (Never mind that the Qualcomm version of the XPS 13 quotes a slightly longer life of up to 27 hours.)

Intel reveals first Lunar Lake laptop CPUs: everything you need to know

What sets Core Ultra 200V apart? Everything.

www.theverge.com

 

[cannedlake240]

which could bring Meteorlake numbers down to what Intel shown

Yeah maybe they picked a wildly inefficient MTL implementation. Still even if LNL is 33% better than 11W MTL, they still need another LNL level uplift to at least get close to apple M series. The problem is that LNL is apparently a one off product and neither intel nor oems appear interested in competing with apple because of the LNL platform cost. Maybe Intel will look at LNL demand and decide to do another generation...

 

[DavidC1]

Yeah maybe they picked a wildly inefficient MTL implementation. Still even if LNL is 33% better than 11W MTL, they still need another LNL level uplift to at least get close to apple M series.

Yes but Intel video shows drop to 5-6W range, not 33% lower, it's nearly 50% lower. That is on Elite's level.

I'm also suspicious of "Idle power". They are using 7W with the screen off really? The Snapdragon uses 4W and 6W? Even the M3 uses 3W?

My 2017 Kabylake Yoga does 3W with screen-on. So my 2017 14nm Intel is superior to Apple M3 right? The problem is both Intel's and NBC's testing methodology is flawed. On battery the power management is better than on AC.

You can see some NBC tests showing "idle" power measurements at 4W, but the notebook is getting 20 hours on a 55WHr battery. Clearly it's way under 4W.

https://www.notebookcheck.net/Sayonara-Alcantara-Microsoft-Surface-Laptop-3-13-5-inch-Core-i5-Review.454000.0.html

-4.4W Idle minimum.
-1.3W Reader(so not 100% idle) battery consumption. 34 hours on a 45W battery. Actually 1.2W since you can't use 100% of battery.

 

[MS_AT]

In Specint they used icc putting others at a disadvantage.

Small correction: ICC is no more. They are now using LLVM derivative. It's called ICX

 

[DavidC1]

@cannedlake240 The package power difference between LNL and M3 might be closer than you think, if my suspicions above are true.

Apple does have an advantage of a tightly integrated system. Apple Intel Macbooks showed top tier battery life too, despite Intel hardware itself sucking in general.

Windows is crazy how much the process use compute for seemingly no reason, all the freaking time. This is why Skymont's power advances are a boon for Windows systems, cause realistically you can't get true idle on Windows. Especially not after 3+ months of using it!

The problem is that LNL is apparently a one off product and neither intel nor oems appear interested in competing with apple because of the LNL platform cost. Maybe Intel will look at LNL demand and decide to do another generation...

They are dimwits driven solely by the greenback that's why.

 

[poke01]

Maximum turbo power will be 37W on Lunar Lake. Of course this is a theoretical maximum as always, but it's a lot lower than on MTL-U with up to 57W. This is good.

Now I want to see the ST power consumption

 

[poke01]

Regardless, Lunar Lake is seriously impressive.

Been telling this since it was rumoured. It’s the best thing from Intel since forever.

 

[cannedlake240]

So my 2017 14nm Intel is superior to Apple M3 right?

Intel laptops can have excellent battery life, even comparable to Apple in some cases, but at the cost of performance. Apple has both industry leading per core perf and stamina

 

[DavidC1]

Now I want to see the ST power consumption

ST power will be the same since they said >20% better performance, but 1/2 the performance at same power, which in tests like Cinebench where it loads the core you'll never see the CPU saying "Oh I'll perform just like the previous generation so I can use half the power.

In order for that to be useful it needs a general purpose workload equivalent of frame rate capping.

Intel laptops can have excellent battery life, even comparable to Apple in some cases, but at the cost of performance. Apple has both industry leading per core perf and stamina

You missed the whole point. Battery life isn't necessarily reflected by the tests and scrutinizing numbers is a waste of time.

And no, current Intel laptops are trash in terms of battery life. They can't even beat their x86 competitor and it spins up fans ridiculously and heats up chassis to do it. It makes Conroe vs Netburst comparisons look like an embarassment.

 

[poke01]

wait it’s not launching this week but the 24th????

That’s like 3 weeks away.

 

[poke01]

https://videocardz.com/newz/intel-unveils-core-ultra-200v-lunar-lake-series-launching-september-24th View attachment 106744
View attachment 106741View attachment 106742View attachment 106743

I love how there is no zero comparison in the Fastest cores slide to M3. Intel compared to the M series before but only when they won.

 

[DavidC1]

I love how there is no zero comparison in the Fastest cores slide to M3. Intel compared to the M series before but only when they won.

Apple M4: Hello?

 

[poke01]

The only M3 comparsion is in this slide and of course Intel used the Intel Complier™.

 

[poke01]

Apple M4: Hello?

That doesn't yet count to Intel as its not in a laptop yet.

 

[DokiDoki]

Cinebench 2024 single-core score puts the 288V bang in the middle between M2 perf (120 pts) and M3 perf (140 pts).

 

[jdubs03]

That doesn't yet count to Intel as its not in a laptop yet.

Neither is the M4 😉

Based on this:

and this:

Performance in spec2017 should thus be somewhere around 8.1. So the M4 is around 32% faster in integer for 1T in the iPad Pro. For the MacBook, one would think it would score a little better.

What is impressive though is what looks like a 29% jump from the 9 185H. Strange though as that’s way more than the 14% IPC increase for LNC vs RWC? Am I missing something here?

 

[poke01]

Neither is the M4 😉

I think you have mistaken me, Intel uses M3 instead of M4 because the latter is not in laptops is what I meant.

Although the definition of ultra slim systems is up to interpretation, I guess Intel means Windows or macOS machines here.

 

[Tigerick]

I have updated the front page with LNL model and specs:

 

[DokiDoki]

I cannot find post-launch SPEC2017 scores for SD X Elite, bar this one which was before launch.

That was running at 3.95 GHz. The SKU Intel compares with is running at 3.4 GHz. So it should score roughly 7.05 at 3.4 GHz instead of 8.19. So with the Intel Compiler, +61% over the SD X Elite puts the 288V at 11.35

That is basically tied with M4 in the iPad Pro, as per Geekerwan.

Obvious caveat is that this is ICC on Intel vs who-knows-what on the SD X Elite.

 

[Saylick]

I have updated the front page with LNL model and specs:

View attachment 106773

TSMC N3B yielding pretty good if the 256V SKU is pretty much fully enabled. Everything above it is just slightly better bins.