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.)

coercitiv · Oct 10, 2024

Magio said:
Is it? N3B for LNL seemed like it allowed for a significant improvement at low power for Intel vs MTL on Intel 4, but on ARL at high power it's not clear where that improvement is compared to RPL on Intel 7...

They are dropping power big time in ST and lightly threaded scenarios. Those of us who asked for more efficiency from Intel on the desktop should keep an open mind here. (it's SoC power but still a very good thing if real)

coercitiv · Oct 10, 2024

Heartbreaker said:
I think you left out the most important slide. Gaming vs 14900K. Note under Title: Same Frame Rates.

This is ZERO percent improvement in gaming performance, according to Intel themselves:

We chewed through this slide yesterday, when it was in chinese

Magio · Oct 10, 2024

511 said:
They did not make mistake for 3nm it was TSMCs lol they bought the first TSMC 3nm Wafer turned out N3 was scrapped N3B was the first one kind of mid for N3 (improved SRAM by few% and slightly dense vs N3E) flavour N3E is the better one in PPA

I'm not saying it's their fault N3B kind of disappointed, but their mistake is to have bet on something that (in my opinion, for ARL-S) isn't worth its cost. Call it unlucky, call it poor planning, still a mistake.

Now their desktop platform is stuck on a dubiously fitting node for 2 years while everyone is going to move to N3E or N3P over the next year. On mobile they have PTL on 18A in the works, but on the desktop they're going to be sitting ducks until NVL.

coercitiv said:
They are dropping power big time in ST and lightly threaded scenarios. Those of us who asked for more efficiency from Intel on the desktop should keep an open mind here. (it's SoC power but still a very good thing if real)

That's fine and I'm not saying there are no benefits to N3B (again, at low power it proved its worth) just that it doesn't seem like it paid out in terms of pure performance, which is still the most important thing for high end desktop chips, considering the cost of those wafers.

Hitman928 · Oct 10, 2024

It’s interesting that Intel slides show that a 285k is on par in gaming to both the 9590x and 14900k. Not the exact same game lineup for each slide but overall impression is equal performance.

511 · Oct 10, 2024

Magio said:
I'm not saying it's their fault N3B kind of disappointed, but their mistake is to have bet on something that (in my opinion, for ARL-S) isn't worth its cost. Call it unlucky, call it poor planning, still a mistake.

Yes one of the reason swan got out N3B is a skip node for many companies outside Intel/Apple

Magio said:
Now their desktop platform is stuck on a dubiously fitting node for 2 years while everyone is going to move to N3E or N3P over the next year. On mobile they have PTL on 18A in the works, but on the desktop they're going to be sitting ducks until NVL.

On Desktop Zen6 will not launch before Q226 so i guess it will be zen5 vs arl launch scenario

511 · Oct 10, 2024

coercitiv said:
They are dropping power big time in ST and lightly threaded scenarios. Those of us who asked for more efficiency from Intel on the desktop should keep an open mind here. (it's SoC power but still a very good thing if real)
View attachment 109103

Gamers will complain but if the reviewers don't blatantly say FPS/watt for CPU is nearly doubled than i will feel I don't what to say there are generational improvement just not where everyone want

511 · Oct 10, 2024

dullard said:
I realize that was a joke post, but I'll run with it as I wait 2 more hours for the official embargo to end. I know we focus a lot on small differences in power here. Less power is better, better for costs, and better for the environment. But for the typical gamer, it really isn't as important as you point out.

The leaked slide shows a geomean average of 73 W less power.

I'll go with the ridiculous assumption of 24/7 gaming with no breaks for sleep.

Lets use the average US energy cost of $0.166 / kWhr.

Lets use the longest delay estimate for Nova Lake that I've seen, Oct 2026. Even though with the cancellation of other CPUs, there are rumors of this date being moved sooner.

0.073 kW * 24 hr use/day * 365 days / year * 2 years * $0.166 / kWhr = $212.31 in those 2 years.

That likely would not buy even the cheapest Nova Lake i5 chip, and certainly a 24/7 gamer would want something better than an i5. And a much more realistic 8 hr gaming per day would be $70.77 over two years.

Euro and US Price difference can be a matter for someone but yes it won't be a humongous save

MoistOintment · Oct 10, 2024

dullard said:
I realize that was a joke post, but I'll run with it as I wait 2 more hours for the official embargo to end. I know we focus a lot on small differences in power here. Less power is better, better for costs, and better for the environment. But for the typical gamer, it really isn't as important as you point out.

The leaked slide shows a geomean average of 73 W less power.

I'll go with the ridiculous assumption of 24/7 gaming with no breaks for sleep.

Lets use the average US energy cost of $0.166 / kWhr.

Lets use the longest delay estimate for Nova Lake that I've seen, Oct 2026. Even though with the cancellation of other CPUs, there are rumors of this date being moved sooner.

0.073 kW * 24 hr use/day * 365 days / year * 2 years * $0.166 / kWhr = $212.31 in those 2 years.

That likely would not buy even the cheapest Nova Lake i5 chip, and certainly a 24/7 gamer would want something better than an i5. And a much more realistic 8 hr gaming per day would be $70.77 over two years.

It's not even just the costs. Heat output into my room is #1 concern to power draw, followed closely in 2nd by fan noise (can't stand hearing my computer, and my PC is in my living room and my wife finds the loud fans annoying), followed 3rd by cooling requirements (don't want a 360 AIO, prefer a smaller case), then followed by power consumption.

Asterox · Oct 10, 2024

Heartbreaker said:
I think you left out the most important slide. Gaming vs 14900K. Note under Title: Same Frame Rates.

This is ZERO percent improvement in gaming performance, according to Intel themselves:
View attachment 109102

I can already see, how the price of the Ryzen 9000 X3D series/for gaming is rising and is not going to drop!

R7 7800X3D same thing, there is no need for price correction + the CPU consumes electricity like two ants!

Hitman928 · Oct 10, 2024

English slides confirm the 285K was running 6400 MT/s memory while the 14900K and AMD CPUs were running 5600 MT/s. Given that and Intel’s “on par” results, I’m guessing ARL comes in behind Zen 5 vanilla in gaming with non-restricted memory.

MarkPost · Oct 10, 2024

Hitman928 said:
English slides confirm the 285K was running 6400 MT/s memory while the 14900K and AMD CPUs were running 5600 MT/s. Given that and Intel’s “on par” results, I’m guessing ARL comes in behind Zen 5 vanilla in gaming with non-restricted memory.

And some ot games with APO for Intel CPU. And dont forget 24h2 and latest AGESA improvements for zen 5. I guess we will see double digits in games in favour of vanilla Zen 5

GTracing · Oct 10, 2024

Hitman928 said:
English slides confirm the 285K was running 6400 MT/s memory while the 14900K and AMD CPUs were running 5600 MT/s. Given that and Intel’s “on par” results, I’m guessing ARL comes in behind Zen 5 vanilla in gaming with non-restricted memory.

Isn't that Jedec vs Jedec? Seems fair to me.

Besides, the sweet spot for RAM overclocking on Zen4 is DDR5-6000, and AFAIK it's similar on Zen5. Raptor Lake can hit DDR5-8000. Arrow Lake be as good or better

Wolverine2349 · Oct 10, 2024

Hitman928 said:
English slides confirm the 285K was running 6400 MT/s memory while the 14900K and AMD CPUs were running 5600 MT/s. Given that and Intel’s “on par” results, I’m guessing ARL comes in behind Zen 5 vanilla in gaming with non-restricted memory.

If that's true that is very disappointing and sad.

I really wanted an on par CPU gaming wise with 7800X3D or at least Raptor Lake that had more than 8 cores and a good in built hardware scheduler and no dual CCD infinity fabric crossing crap.

Maybe raising power limits similar to Raptor Lake levels will make it better.

The selling point seems better power efficiency.

I wonder if it can perform belter than Raptor Lake and on par with 7800X3D if power limits are removed, but this time with strong platinum stability and no degradation problems. Well power limits removed to an extent. Obviously do not want 400W or even over 300W like Raptor Lake, but the 250W or maybe up to 300W CPU only.

gdansk · Oct 10, 2024

I was looking through the footnotes (here) but did not find the relevant details.
Did Intel disclose which compiler was used for SPECrate 2017 comparisons somewhere else?

MarkPost · Oct 10, 2024

GTracing said:
Isn't that Jedec vs Jedec? Seems fair to me.

Besides, the sweet spot for RAM overclocking on Zen4 is DDR5-6000, and AFAIK it's similar on Zen5. Raptor Lake can hit DDR5-8000. Arrow Lake be as good or better

High memory speed is just a marketing thing. We all are tired to see benches where ~8000 DDR5 speed just means a barely 3% improvement in gaming (and I am talking about RPL. Dont see any reason it will be different with ARL)

dullard · Oct 10, 2024

S

gdansk said:
I was looking through the footnotes (here) but did not find the relevant details.
Did Intel disclose which compiler was used for SPECrate 2017 comparisons somewhere else?

Seems like Intel needs to revise that table. They have a math error in there, incrementing the date by one year each line. Just thought it was humorous.

Hitman928 · Oct 10, 2024

GTracing said:
Isn't that Jedec vs Jedec? Seems fair to me.

Besides, the sweet spot for RAM overclocking on Zen4 is DDR5-6000, and AFAIK it's similar on Zen5. Raptor Lake can hit DDR5-8000. Arrow Lake be as good or better

Gamers don't run JEDEC.

6000 MT/s is a sweet pot for Zen in terms of perf/$ and early on compatibility. Zen 4 and Zen 5 can hit higher speeds pretty easily now, though going over 6200/6400 MT/s usually isn't helpful as you have to switch to a lower fabric clock ratio, but going from 5600 to 6200/6400 MT/s gives significant performance gains and is not difficult anymore with Zen 5. If you really want to get high speeds, Zen 5 can get up to 7800/8000 MT/s as well, but again, it doesn't typically make sense to do so as you are spending more money on faster memory for no real gain since you have to drop the clock ratio.

Even on RPL, you plateau in gaming performance pretty hard somewhere in the 6800 - 7200 MT/s range, but again, going from 5600 MT/s to even 6400 MT/s gives good performance gains, though Zen typically responds a little better to faster memory (assuming good timings for lowered latency).

MS_AT · Oct 10, 2024

gdansk said:
I was looking through the footnotes (here) but did not find the relevant details.
Did Intel disclose which compiler was used for SPECrate 2017 comparisons somewhere else?

Some slides are quoting intel compiler [those showing skymont improvements vs crestmont] others are silent about it

Hitman928 · Oct 10, 2024

Wolverine2349 said:
If that's true that is very disappointing and sad.

I really wanted an on par CPU gaming wise with 7800X3D or at least Raptor Lake that had more than 8 cores and a good in built hardware scheduler and no dual CCD infinity fabric crossing crap.

Maybe raising power limits similar to Raptor Lake levels will make it better.

The selling point seems better power efficiency.

I wonder if it can perform belter than Raptor Lake and on par with 7800X3D if power limits are removed, but this time with strong platinum stability and no degradation problems. Well power limits removed to an extent. Obviously do not want 400W or even over 300W like Raptor Lake, but the 250W or maybe up to 300W CPU only.

From what I understand, the numbers Intel showed were for PL1=PL2=250 W for the 285K except for one or two slides that talked about efficiency at 125 W.

Edit: Yep, PL1=PL2= 250 W for Intel performance numbers. Thanks to @gdansk for the footnotes link.

Edit 2: Looks like Intel is using W11 24H2 at least. I'm assuming the build is adequate for the AMD performance fix, can anyone confirm?

jdubs03 · Oct 10, 2024

Not sure this has been posted, but here is the full slide deck for ARL: https://download.intel.com/newsroom...Intel-Core-Ultra-200S-Series-Presentation.pdf

dttprofessor · Oct 10, 2024

ddr5 6400

dullard · Oct 10, 2024

Wolverine2349 said:
I wonder if it can perform belter than Raptor Lake and on par with 7800X3D if power limits are removed, but this time with strong platinum stability and no degradation problems. Well power limits removed to an extent. Obviously do not want 400W or even over 300W like Raptor Lake, but the 250W or maybe up to 300W CPU only.

Intel's slides show that gaming performance is linear with power for the 285K (same performance per watt, at least when going from 125 W to 250 W). So, if that slide is correct, then 250 W power is double the gaming performance as 125 W. If that holds at higher power levels, then 300 W would have about a 20% gaming boost over 250 W.

GTracing · Oct 10, 2024

Hitman928 said:
Gamers don't run JEDEC.

6000 MT/s is a sweet pot for Zen in terms of perf/$ and early on compatibility. Zen 4 and Zen 5 can hit higher speeds pretty easily now, though going over 6200/6400 MT/s usually isn't helpful as you have to switch to a lower fabric clock ratio, but going from 5600 to 6200/6400 MT/s gives significant performance gains and is not difficult anymore with Zen 5. If you really want to get high speeds, Zen 5 can get up to 7800/8000 MT/s as well, but again, it doesn't typically make sense to do so as you are spending more money on faster memory for no real gain since you have to drop the clock ratio.

Even on RPL, you plateau in gaming performance pretty hard somewhere in the 6800 - 7200 MT/s range, but again, going from 5600 MT/s to even 6400 MT/s gives good performance gains, though Zen typically responds a little better to faster memory (assuming good timings for lowered latency).

In my experience from talking to my PC gamer friends in real life, the majority of them don't overclock their RAM. Enthusiasts do, but not the average gamer.

Fair point for Ryzen gaining more from going to 6400 than intel does going to 8000.

MarkPost · Oct 10, 2024

Looking at footnotes, it seems they already used 24H2 and BIOS with new AGESA. But APO enabled remains.

btw according to the footnotes 285K has 4% lower power consumption that 14900K running CB 24 MT as well as general MT tasks. It means PL2 250W is reached.

Intel Core Ultra 200S series processors

Power Efficiency

Up to 4% lower processor power while running Cinebench 2024 Multi Core

Intel® Core™ Ultra 9 285K

Processor: Intel® Core™ Ultra 9 285K Processor, 24C24T (8P + 16E); PL1 = PL2 =250W; Memory: 2x16GB DDR5-6400MHz; Storage: Samsung 980 Pro 500GB; Display Resolution: 1920x1080; OS: Microsoft Windows 11 Pro 24H2 OS Build 26100.1457; Integrated Graphics: Intel® Graphics; Integrated Graphics Driver: 32.0.101.5866; Motherboard: Intel RVP; BIOS Version: MTLSFW1.R00.4165.D02.2404300646; NPU Driver: 32.0.100.2540; Power Plan set to High Performance; Resizable BAR: On; Trusted Platform Module: On; VBS: On; Defender: On Processor: Intel® Core™ i9-14900K Processor, 24C32T (8P + 16E); PL1 = PL2 =253W; Memory: 2x16GB DDR5-5600MHz; Storage: Samsung 980 Pro 500GB; Display Resolution: 1920x1080; OS: Microsoft Windows 11 Pro 24H2 OS Build 26100.1457; Integrated Graphics: Intel® UHD Graphics; Integrated Graphics Driver: 31.0.101.5537; Motherboard: Intel RVP; BIOS Version: RPLSFWI1.R00.5045.A00.2401260733; Power Plan set to High Performance; Resizable BAR: On; Trusted Platform Module: On; VBS: On; Defender: On

As measured by average processor power while running Cinebench 2024 Multi Core benchmark

As of October 2029

Intel Core Ultra 200S series processors

Power Efficiency

Up to 4% lower processor power during multi threaded tasks

Intel® Core™ Ultra 9 285K

Processor: Intel® Core™ Ultra 9 285K Processor, 24C24T (8P + 16E); PL1 = PL2 =250W; Memory: 2x16GB DDR5-6400MHz; Storage: Samsung 980 Pro 500GB; Display Resolution: 1920x1080; OS: Microsoft Windows 11 Pro 24H2 OS Build 26100.1457; Integrated Graphics: Intel® Graphics; Integrated Graphics Driver: 32.0.101.5866; Motherboard: Intel RVP; BIOS Version: MTLSFW1.R00.4165.D02.2404300646; NPU Driver: 32.0.100.2540; Power Plan set to High Performance; Resizable BAR: On; Trusted Platform Module: On; VBS: On; Defender: On Processor: Intel® Core™ i9-14900K Processor, 24C32T (8P + 16E); PL1 = PL2 =253W; Memory: 2x16GB DDR5-5600MHz; Storage: Samsung 980 Pro 500GB; Display Resolution: 1920x1080; OS: Microsoft Windows 11 Pro 24H2 OS Build 26100.1457; Integrated Graphics: Intel® UHD Graphics; Integrated Graphics Driver: 31.0.101.5537; Motherboard: Intel RVP; BIOS Version: RPLSFWI1.R00.5045.A00.2401260733; Power Plan set to High Performance; Resizable BAR: On; Trusted Platform Module: On; VBS: On; Defender: On

As measured by average processor power while running Cinebench 2024 Multi Core benchmark

As of October 2031

Hitman928 · Oct 10, 2024

Edit: I was mistaken, I read the wrong footnote for this slide.

This slide from Intel is silly. Why? Because to get these power numbers, they set PL1=125 W for the 285K and PL1 = 253 W for the 14900K. Yes, they say performance is on par and clearly the 285K is more efficient as that won't effect the 1t results, but for the others that are lightly threaded stuff, if you dropped the PL1 to 125 W for the 14900K, you wouldn't lose very much performance at all. Why not just set both to 125 W and show a like for like comparison? The 285K would still look good, just maybe not quite as good, but at least it would be an honest comparison.

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

Senior member

Attachments

Diamond Member

Diamond Member

Member

Diamond Member

Golden Member

Golden Member

Golden Member

Member

Golden Member

Diamond Member

Senior member

Member

Senior member

Diamond Member

Senior member

Elite Member

Diamond Member

Senior member

Diamond Member

Golden Member

Member

Elite Member

Member

Senior member

Diamond Member