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 (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.)

 

[Hitman928]

 

[Josh128]

Fastest iGPU?
View attachment 108127View attachment 108128

Second fastest, loses to Strix at high and ultra, which are more GPU intensive. 258V winning at low indicates faster CPU ST perf.

 

[Hitman928]

This is just int, still waiting on fp. Need both to determine approx IPC.

SPECint has been the standard for IPC for quite a while here.

 

[Hitman928]

Battery life and GPU improvements seem to be the shining points, though we need a lot more samples to get a firm grasp on how good battery life really is. More GPU testing as well, it will be interesting to see if Intel was able to fix the areas were GPU performance choked previously.

 

[Josh128]

SPECint has been the standard for IPC for quite a while here.

SPECfp doesnt matter?

 

[Hitman928]

SPECfp doesnt matter?

Not much. It more or less gets ignored. It's actually not a great test for FP performance.

 

[poke01]

This leads be to believe that N3B was a bad node. Both for Apple and Intel

View attachment 108131

Battery life and GPU improvements seem to be the shining points, though we need a lot more samples to get a firm grasp on how good battery life really is. More GPU testing as well, it will be interesting to see if Intel was able to fix the areas were GPU performance choked previously.

how does it lose to a fanless Air? Talking MT obviously

 

[cebri1]

Second fastest, loses to Strix at high and ultra, which are more GPU intensive. 258V winning at low indicates faster CPU ST perf.

Keep in mind this GPU doesn’t boost to 2.1Ghz like the 140V in the 288V.

 

[igor_kavinski]

Do you expect IPC to be much higher for ARL? I'm not counting on a big uplift, it's the same core with some more LLC.

Might be neck and neck with Zen 5 in integer subtests or barely be better overall in INT but may lose in SpecFP.

 

[Hitman928]

Might be neck and neck with Zen 5 in integer subtests or barely be better overall in INT but may lose in SpecFP.

I agree with you, but the general reaction to Zen 5 was that it was a huge disappointment. . .

 

[511]

Which freq for which core?

3.7 for 258V and 3.3 for 5C

 

[cebri1]

View attachment 108131

Battery life and GPU improvements seem to be the shining points, though we need a lot more samples to get a firm grasp on how good battery life really is. More GPU testing as well, it will be interesting to see if Intel was able to fix the areas were GPU performance choked previously.

Notebook check get 120 points in ST, which puts it 6% ahead of Strix. That’s another shinning point in my book.

 

[Hitman928]

This leads be to believe that N3B was a bad node. Both for Apple and Intel

how does it lose to a fanless Air? Talking MT obviously

Apple cores are still king in terms of efficiency.

 

[511]

Notebook check get 120 points in ST, which puts it 6% ahead of Strix. That’s another shinning point in my book.

While consuming way less power lol

 

[cebri1]

:O

 

[tsamolotoff]

Some results from Gordon Ung

 

[511]

Second fastest, loses to Strix at high and ultra, which are more GPU intensive. 258V winning at low indicates faster CPU ST perf.

We need higher power limits for both lol

 

[igor_kavinski]

I agree with you, but the general reaction to Zen 5 was that it was a huge disappointment. . .

Arrow Lake would be a better disappointment than degrading Raptor Lakes

At least users won't have to worry anymore. TSMC doesn't have a history of bad process issues. Yet.

 

[Josh128]

The ST efficiency beats the hell out of Strix, but this is a 4.8 vs 5.1 scenario, keep that in mind. Downclocking Strix to 4.8 may boost its efficiency a decent amount. What I find more intriguing is that the Lion Cove R23 ST is 7% behind Strix (4.8 vs 5.1), and is even about 0.75% behind when clock normalized. Just going off of this, Lion Cove is essentially identical to Zen 5 in R23 IPC. That doesnt bode well for those predicting 2500+ R23 ST for the 285K. Unless this Zenbook S14 is not holding its ST boost well during the test. In R15 and R20 ST, its behind 18% and 23% respectively. Ironically, Lunar Lake beats Strix in R24 ST by around 12 to 13% clock normalized. Wonder why such a huge the role reversal vs R15, 20, and 23?

 

[cebri1]

The ST efficiency beats the hell out of Strix, but this is a 4.8 vs 5.1 scenario, keep that in mind. Downclocking Strix to 4.8 may boost its efficiency a decent amount. What I find more intriguing is that the Lion Cove R23 ST is 7% behind Strix (4.8 vs 5.1), and is even about 0.75% behind when clock normalized. Just going off of this, Lion Cove is essentially identical to Zen 5 in R23 IPC. That doesnt bode well for those predicting 2500+ R23 ST for the 285K. Unless this Zenbook S14 is not holding its ST boost well during the test. In R15 and R20 ST, its behind 18% and 23% respectively. Ironically, Lunar Lake beats Strix in R24 ST by around 12 to 13% clock normalized. Wonder why such a huge the role reversal vs R15, 20, and 23?


View attachment 108134

In CB24 beats it by 6% even with the clock advantage, as you say 13% normalized. So I would wait for ARL and see.

 

[Hitman928]

The ST efficiency beats the hell out of Strix, but this is a 4.8 vs 5.1 scenario, keep that in mind. Downclocking Strix to 4.8 may boost its efficiency a decent amount. What I find more intriguing is that the Lion Cove R23 ST is 7% behind Strix (4.8 vs 5.1), and is even about 0.75% behind when clock normalized. Just going off of this, Lion Cove is essentially identical to Zen 5 in R23 IPC. That doesnt bode well for those predicting 2500+ R23 ST for the 285K. Unless this Zenbook S14 is not holding its ST boost well during the test. In R15 and R20 ST, its behind 18% and 23% respectively. Ironically, Lunar Lake beats Strix in R24 ST by around 12 to 13% clock normalized. Wonder why such a huge the role reversal vs R15, 20, and 23?


View attachment 108134

CB24 has some sensitivity to memory bandwidth where the previous versions didn’t.

 

[Hitman928]

View attachment 108135

In CB24 beats it by 6% even with the clock advantage, as you say 13% normalized. So I would wait for ARL and see.

HX370 at proper boost and high speed memory scores about 120 in CB24 ST, so LNL has about a 5% advantage there.

 

[9949asd]

HX370 at proper boost and high speed memory scores about 120 in CB24 ST, so LNL has about a 5% advantage there.

how about max out 288v st will be 130？

 

[cebri1]

pretty impressive generational jump. Arc team must be very pleased.

 

[Josh128]

Interestingly, for battery life, the Strix HX 370 based Vivobook S 14 still beats everything, including this Lunar Lake Zenbook S14. Display obviously has a tremendous amount of influence on battery life, as the lower resolution seems to suck much less power at the same screen size than the UX5406.

Would be very interesting to see how a Vivobook version of Lunar Lake with same display compares.