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	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.)

mikk · Mar 2, 2024

Any experienced tester uses HWinfo, Core Temp or whatever to make sure it really runs with the desired power. You would expect they did this.

Markfw · Mar 2, 2024

mikk said:
Any experienced tester uses HWinfo, Core Temp or whatever to make sure it really runs with the desired power. You would expect they did this.

I would certainly not expect that. Until I see multiple benchmarks that I can read (read NOT chinese) that agree, I will not believe any of these.

H433x0n · Mar 2, 2024

Hitman928 said:
Using 3rd party tools on a laptop for power/frequency/voltage manipulation is never a guarantee. Just because it works on desktop or on a different laptop, doesn't mean it will work correctly on every model, I thought this was common knowledge here? As far as I know, he doesn't even show how he was setting PL1 and PL2 numbers. So yes, most likely the setting of PL2=PL1 was not being enforced correctly and the reviewer didn't do a sanity check to make sure it was working as presented. The numbers bear this out as I've shown.

I'm not sure how you can see the chart I posted and think it looks reasonable. If the chip was heat soaked at 70/95 W, how did it continue to increase in performance up to 95/95 W? It is literally impossible to get the results they did and for the chart to be with PL2=PL1. I will gladly eat crow on this if you can show me any other example of a processor that both plateaus in performance with increasing power, and increases in performance with increasing power simultaneously. If you can find that mythical CPU, I will take back my whole argument.

Edit: he does show he is using the Armoury Crate software to control the LEDs on the laptop, so presumably he is using it to set PL1 and PL2 as well, but again, that doesn't mean it's actually working as expected.

This just reinforces my opinion that using laptop reviews to determine CPU performance/watt is a complete waste of time. I thought that maybe this one would’ve been a useful comparison since it had both chips using the same chassis but nope.. the results were too close (which is clearly not physically possible!) therefore the data must be invalid.

tamz_msc · Mar 2, 2024

H433x0n said:
This just reinforces my opinion that using laptop reviews to determine CPU performance/watt is a complete waste of time. I thought that maybe this one would’ve been a useful comparison since it had both chips using the same chassis but nope.. the results were too close (which is clearly not physically possible!) therefore the data must be invalid.

Very few "reviewers" know how to how to benchmark mobile Intel CPUs. The first thing you have to do to make any sensible power measurements using HWinfo is to use throttlestop to disable DTT.

I see no evidence that people actually do that, or at the very least, establish that dynamic tuning is disabled during the benchmark run.

tamz_msc · Mar 2, 2024

Here is the difference DTT makes in practice:

i7-11370H PL2= 51 W, PL1= 30 W

DTT OFF:

DTT ON:

Sure, performance is a bit lower, but you get similar efficiency and energy consumption vs. DTT off when blasting an all-core load like CB R23 - DTT functions primarily to give better user experience in terms of heat and noise.

DTT ON:

DTT OFF:

DTT in case of a light variable workload - which is Geekbench 5 - actually sacrifices a bit of efficiency for a higher score (as evidenced by the run times).

So efficiency measurement in reviews is completely dependent on context - no one single number can express it when there are so many variables at play in addition to power limits.

Hulk · Mar 2, 2024

Meteor Lake is a little better than Raptor Lake mobile it appears. Kind of blah overall for me because Intel set us up to have our socks knocked off.
On the other hand, the fact that they got all those tiles, Foveros, NPU, etc... working is kind of impressive.
Hopefully we'll see the performance end of things with Arrow and Lunar Lake because AMD is definitely going to up the ante with Zen 5.
When I go for my next build I'm going to have to take a hard look at Zen 5 and ARL. Might be time for my first AMD build!

Henry swagger · Mar 3, 2024

Hulk said:
Meteor Lake is a little better than Raptor Lake mobile it appears. Kind of blah overall for me because Intel set us up to have our socks knocked off.
On the other hand, the fact that they got all those tiles, Foveros, NPU, etc... working is kind of impressive.
Hopefully we'll see the performance end of things with Arrow and Lunar Lake because AMD is definitely going to up the ante with Zen 5.
When I go for my next build I'm going to have to take a hard look at Zen 5 and ARL. Might be time for my first AMD build!

Intel will sell millions of meteor lake.. while amd struggle to get vendors.. i think intel are happy

igor_kavinski · Mar 3, 2024

Pat (to his chief foundry engineer): Can we produce more 185H CPUs?

Engineer: Unfortunately, the process isn't that refined yet. We need more time.

Pat: Any other way you can think of?

Engineer: We produce more wafers and the chance of finding 185H quality chips in them increase proportionately.

Pat: "facepalm"

Hulk · Mar 3, 2024

Henry swagger said:
Intel will sell millions of meteor lake.. while amd struggle to get vendors.. i think intel are happy

No doubt. But every near miss adds up. Ask IBM.

How IBM Crashed, Then Rebooted With a $24 Billion Comeback

IBM was once on top of the world. A series of mistakes brought them to the brink of disaster. Learn how they turned their fortunes around.

leaders.com

FlameTail · Mar 3, 2024

Henry swagger said:
Intel will sell millions of meteor lake.. while amd struggle to get vendors.. i think intel are happy

Meanwhile Qualcomm lands a spectacular kick from behind.

cortexa99 · Mar 3, 2024

北森瓦版 - 【怪情報】“Arrow Lake-S”―Core i9 15900K, i7 15700K（？）のスペック

北森瓦版 - Northwood Blog (Author : 北森八雲. Since July 10, 2006.)

northwood.blog.fc2.com

It seems Yuuki_Ans posted ARL-S details and deleted at once, but unfortunately some lurkers already had a backup before he deleted the post

◇Core i9 15900K
　・コア・スレッド数：24-core/24-thread（？？）
　・Performance Core：8-core, Max 5.80GHz？
　・Efficent Core：16-core, Max 4.40GHz
　・L2 cache：2.5MB×8 (P-core) + 4MB×4 (E-core) = Total 36MB
　・L3 cache：36MB (= 24MB (3MB×8-core) + 12MB (3MB×4 cluster))
　・PL1 125W / PL2 177W
　・Memory：2-channel DDR5-6400
　・GPU：Xe-LPG 32 EU / 2 Xe-core（？）
　・NPU：10 TOPS
　・Socket：LGA1851
　・I/O：PCI-Express 5.0×20 (+ PCI-Express 4.0×4？)

◇Core i7 15700K
　・コア・スレッド数：20-core/20-thread
　・Performance Core：8-core, Max 5.50GHz？
　・Efficent Core：12-core, Max 4.20GHz
　・L2 cache：2.5MB×8 (P-core) + 4MB×3 (E-core) Total 32MB
　・L3 cache：36MB (= 24MB (3MB×8-core) + 9MB (3MB×3 cluster))
　・PL1 125W / PL2 177W
　・Memory：2-channel DDR5-6400
　・GPU：Xe-LPG 32 EU / 2 Xe-core（？）
　・NPU：10 TOPS
　・Socket：LGA1851
　・I/O：PCI-Express 5.0×20 (+ PCI-Express 4.0×4？)

So ARL-S reach 5.8Ghz turbo after all?

EDIT: Yuuki_AnS just clarify it's fake.

tamz_msc · Mar 3, 2024

Shouldn't the Core-iX branding be gone by now?

mikk · Mar 3, 2024

cortexa99 said:
北森瓦版 - 【怪情報】“Arrow Lake-S”―Core i9 15900K, i7 15700K（？）のスペック

北森瓦版 - Northwood Blog (Author : 北森八雲. Since July 10, 2006.)

northwood.blog.fc2.com

It seems Yuuki_Ans posted ARL-S details and deleted at once, but unfortunately some lurkers already had a backup before he deleted the post

So ARL-S reach 5.8Ghz turbo after all?

Xe-LPG 32EU is wrong it gets 4 Xe cores with 64 Vector engines. I would say it's too early for final clock speeds (why is there a ?), same for the final naming. Is there no screenshot from the "leak"? It doesn't look credible to me.

ondma · Mar 3, 2024

cortexa99 said:
北森瓦版 - 【怪情報】“Arrow Lake-S”―Core i9 15900K, i7 15700K（？）のスペック

北森瓦版 - Northwood Blog (Author : 北森八雲. Since July 10, 2006.)

northwood.blog.fc2.com

It seems Yuuki_Ans posted ARL-S details and deleted at once, but unfortunately some lurkers already had a backup before he deleted the post

So ARL-S reach 5.8Ghz turbo after all?

There is a question mark after that 5.8 ghz claim, so I would take it with a very large grain of salt. I just dont see how ARL is going to be competitive without hyperthreading or an increase in E cores.

Edit: it also says the CPU core will be on 20A, not a TSMC node.

H433x0n · Mar 3, 2024

cortexa99 said:
北森瓦版 - 【怪情報】“Arrow Lake-S”―Core i9 15900K, i7 15700K（？）のスペック

北森瓦版 - Northwood Blog (Author : 北森八雲. Since July 10, 2006.)

northwood.blog.fc2.com

It seems Yuuki_Ans posted ARL-S details and deleted at once, but unfortunately some lurkers already had a backup before he deleted the post

So ARL-S reach 5.8Ghz turbo after all?

Some of this data I know to be outdated / wrong (for example the PL2). I don't think the top clocks will end up at 5.8ghz/4.4ghz for P & E cores respectively. It will have better clocks than what was projected 6+ months ago but 5.8ghz is probably still a stretch. The process being on 20A is just hopium, the SKUs that matter will be on N3B 🤮

cortexa99 · Mar 3, 2024

Ouch, well, after reconsideration I guess there may be something wrong, the source "ithome.com" that japanese author 北森八雲 cited is the witnesses of Yuuki_AnS's twitter message, but the site ithome.com didn't provide any screenshot of that, also the ARL-S spec and naming scheme are fishy. It's either ithome read something wrongly or brought up old data and made them into a "news", even the japanese author has confused by these weird leak.

DrMrLordX · Mar 3, 2024

Henry swagger said:
while amd struggle to get vendors

AMD isn't struggling to "get vendors".

Markfw · Mar 3, 2024

DrMrLordX said:
AMD isn't struggling to "get vendors".

Henry is struggling to accept the fact that CURRENTLY AMD is in the lead in every area.....

Hulk · Mar 3, 2024

Markfw said:
Henry is struggling to accept the fact that CURRENTLY AMD is in the lead in every area.....

I've always been an Intel person and still am but I have to agree with this. AMD is right there in desktop performance, better in efficiency, in the lead in mobile, and I think the server space as well. The mobile thing is weird. Seems like AMD is better but I don't see their parts in a lot of systems. I wonder if Intel still has leverage with manufacturers?

If it was a running race I'd say it's close but AMD has the lead. AMD only needs to continue at the pace is it currently running while Intel needs a burst of speed to catch up.

Honestly I love this neck-and-neck competition we've had for the past few years between Intel and AMD as it's good for us for things to discuss and just plain fun to see both sides pushed.

DrMrLordX · Mar 3, 2024

Hulk said:
I wonder if Intel still has leverage with manufacturers?

It's a big question mark, but I still strongly suspect that early runs of AMD mobile SoCs wind up in ODM builds that will never go on sale to the public. That being said, I haven't seen the evidence to conclusively prove it. AMD's revenue sheets aren't really providing enough information to prove/disprove it. All we can say for sure is that AMD mobile SoC's exist and that AMD sells them to someone, in some quantity.

Intel does have continuing relationships with OEMs, but to what extent their backroom shenanigans play a factor is unknown. All that can be said for sure is that Intel hasn't had major problems providing 10nm/Intel 7 volume to OEMs. Their Intel 4 volume has not been as impressive to date, and in the end Meteor Lake probably won't ship as many units as Raptor Lake mobile.

tamz_msc · Mar 3, 2024

AMD Client operating margin - 4%
Intel CCG operating margin - 33%.

That should tell you all you need to know about how AMD fares at present, even when technically* ahead in the desktop and mobile space.

*with lots of caveats.

H433x0n · Mar 3, 2024

Hulk said:
I've always been an Intel person and still am but I have to agree with this. AMD is right there in desktop performance, better in efficiency, in the lead in mobile, and I think the server space as well. The mobile thing is weird. Seems like AMD is better but I don't see their parts in a lot of systems. I wonder if Intel still has leverage with manufacturers?

If it was a running race I'd say it's close but AMD has the lead. AMD only needs to continue at the pace is it currently running while Intel needs a burst of speed to catch up.

Honestly I love this neck-and-neck competition we've had for the past few years between Intel and AMD as it's good for us for things to discuss and just plain fun to see both sides pushed.

The stakes for Zen 5 and Lion Cove can’t be overstated. I would argue that the upcoming generation will determine market share for the next 3 years. It comes down to whether or not Zen 5 delivers the >20% IPC increase, if it does then Intel is arguably in a worse spot than it is now.

Goop_reformed · Mar 3, 2024

tamz_msc said:
AMD Client operating margin - 4%
Intel CCG operating margin - 33%.

That should tell you all you need to know about how AMD fares at present, even when technically* ahead in the desktop and mobile space.

*with lots of caveats.

Now post the important stuff

AMDK11 · Mar 4, 2024

Hulk said:
AMD was doing pretty well from a technological point-of-view up until Conroe, which of course is why Intel made a 180 from Netburst to Core. Conroe was monumental, nearly doubling the IPC of the last iteration of the P4. In fact, it put them so far ahead of AMD that they coasted on it for years and that is part of the reason they are struggling today. Don't underestimate Conroe because it didn't have any competition. It didn't have any competition because it was that good.

It is true that Conroe is a large development of the core, somewhere around 40%. It was the introduction of several new techniques and improvement of those known from Baniasz, Yonaha and Netbrust. Nevertheless, IPC was higher than Yonah and K8 on average +10-20%. Is it monumental? Not compared to SunnyCove or GoldenCove which are comparable to Conroe in terms of core logic expansion and IPC increase.

Compared to Netbrust, the IPC was much higher, but the very low IPC of Netbrust (Pentium 4) was due to the microarchitecture itself, which only provided a 1-way decoder. Netbrust has an IPC lower than Yonah, Banias and even Pentium III.

Additionally, the impression of Conroe's monumentality was intensified by the fact that up to Core 2 IPC and the resulting expansion of cores were very small because performance was increased mainly by increasing the clock speed.

Aapje · Mar 4, 2024

AMDK11 said:
Compared to Netbrust, the IPC was much higher, but the very low IPC of Netbrust (Pentium 4) was due to the microarchitecture itself, which only provided a 1-way decoder. Netbrust has an IPC lower than Yonah, Banias and even Pentium III.

The plan with NetBurst was to achieve 10 Ghz, with a low IPC design that could be run at very high clock speeds. They couldn't get the power usage/heat under control though, so Core was the pivot to a design with much higher IPC, designed to work with lower clock speeds.

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

Senior member

Attachments

Diamond Member

Moderator Emeritus, Elite Member

Senior member

Diamond Member

Diamond Member

Diamond Member

Senior member

Lifer

Diamond Member

Platinum Member

Senior member

Diamond Member

Diamond Member

Platinum Member

Senior member

Senior member

Lifer

Moderator Emeritus, Elite Member

Diamond Member

Lifer

Diamond Member

Senior member

Member

Senior member

Golden Member