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.)

 

[witeken]

People like @uzzi38 who subscribe to the "Meteor Lake GPU was always 5nm" forget that besides the N3 slide at Investor Meeting, there have been rumors, from multiple sources, that Intel would be first in line for N3 (H2'22 production) as far back as 2020.

Those rumors were confirmed again in 2021. And again in 2022.

Intel Places Multi-Billion-Dollar Wafer Order At TSMC, Murthy Gone (NASDAQ:INTC)

On the verge of Intel's 7nm delay, rumors have surfaced of a multi-dollar order at TSMC for 6N wafers and 5N for Ponte Vecchio.

seekingalpha.com

 

[uzzi38]

People like @uzzi38 who subscribe to the "Meteor Lake GPU was always 5nm" forget that besides the N3 slide at Investor Meeting, there have been rumors, from multiple sources, that Intel would be first in line for N3 (H2'22 production) as far back as 2020.

Those rumors were confirmed again in 2021. And again in 2022.

Intel Places Multi-Billion-Dollar Wafer Order At TSMC, Murthy Gone (NASDAQ:INTC)

On the verge of Intel's 7nm delay, rumors have surfaced of a multi-dollar order at TSMC for 6N wafers and 5N for Ponte Vecchio.

seekingalpha.com

Such rumours are entirely meaningless. Where were they actually confirmed in thsoe articles btw? One of those is a dead link, the other is claiming that Intel would use N3 for it's i3 lineup in 2022, and N5 in 2021 (like that turned out well, right?).

Come on.

 

[lightisgood]

People like @uzzi38 who subscribe to the "Meteor Lake GPU was always 5nm" forget that besides the N3 slide at Investor Meeting, there have been rumors, from multiple sources, that Intel would be first in line for N3 (H2'22 production) as far back as 2020.

Those rumors were confirmed again in 2021. And again in 2022.

Intel Places Multi-Billion-Dollar Wafer Order At TSMC, Murthy Gone (NASDAQ:INTC)

On the verge of Intel's 7nm delay, rumors have surfaced of a multi-dollar order at TSMC for 6N wafers and 5N for Ponte Vecchio.

seekingalpha.com

Frankly speaking, TSMC N3 is rotten process.

Intel must have revised first plan of MTL and reduced MTL's competitiveness.
However, IMO, rotten N3 could give Intel 4/3 process famiry much larger benefit.

This is recent article of N3.

Apple: First To Use TSMC’s '3 Nm' Process (NASDAQ:AAPL)

Taiwan’s Commercial Times reports that Apple will be the first customer for TSMC’s 3 nm process in the second half of the year. See what this means for AAPL.

seekingalpha.com

 

[jpiniero]

Frankly speaking, TSMC N3 is rotten process.

There doesn't seem to be anything wrong with it now. It's just very overpriced. That was the strange part about using N3 - a $40 IGP chiplet makes zero sense.

The one thing I do wonder about is that if they are really using 22FFL for the base tile, where they would get the capacity. I imagine they probably still have the tools but they would need a lot of space.

 

[DrMrLordX]

All of the things shared by Intel thus far have been either with either a 2+8 or a 6+8 die afaik.

Indeed. If they have anything else planned, it somehow has flown under all the radars of various leakers.

 

[eek2121]

Frankly speaking, TSMC N3 is rotten process.

Intel must have revised first plan of MTL and reduced MTL's competitiveness.
However, IMO, rotten N3 could give Intel 4/3 process famiry much larger benefit.

This is recent article of N3.

Apple: First To Use TSMC’s '3 Nm' Process (NASDAQ:AAPL)

Taiwan’s Commercial Times reports that Apple will be the first customer for TSMC’s 3 nm process in the second half of the year. See what this means for AAPL.

seekingalpha.com

It. was. never. planned.

There are existing slides from Intel themselves long before the N3 rumor exploded. Intel communicated to the press that they would be using N6/N5 long before the rumor exploded.

You were misled by folks trying to drive up subscriber counts and views to videos and websites.

Indeed. If they have anything else planned, it somehow has flown under all the radars of various leakers.

Do we know how much of an IPC uplift Intel plans to bring for both cores? Maybe they don’t need more cores. Otherwise I could see them simply adding a second CPU die to make 12+16.

 

[DrMrLordX]

Do we know how much of an IPC uplift Intel plans to bring for both cores?

The Redwood Cove is apparently not a radical change from Golden/Raptor Cove. Supposedly. We'll see more next year perhaps. As for the new -mont? Uh, Crestmont? Complete unknown.

Otherwise I could see them simply adding a second CPU die to make 12+16.

I've mused on that as well. Until we see desktop packages we won't know if Intel will try something like that. The 6+8 tile is pretty small . . .

 

[Kaluan]

IMHO if the slide would have also said "External N5" than maybe a lot of this rumour mess would've never existed.

 

[dullard]

People like @uzzi38 who subscribe to the "Meteor Lake GPU was always 5nm" forget that besides the N3 slide at Investor Meeting, there have been rumors, from multiple sources, that Intel would be first in line for N3 (H2'22 production) as far back as 2020.

Those rumors were confirmed again in 2021. And again in 2022.

Intel Places Multi-Billion-Dollar Wafer Order At TSMC, Murthy Gone (NASDAQ:INTC)

On the verge of Intel's 7nm delay, rumors have surfaced of a multi-dollar order at TSMC for 6N wafers and 5N for Ponte Vecchio.

seekingalpha.com

Psst: Reread your links as Arrow Lake. Those links do not mention Meteor Lake (although, the 2021 link doesn't work, so I didn't read it).

In order to have Arrow Lake in late 2023/early 2024, Intel needs N3 up and running in 2022 for engineering and qualification samples. Then, once they pass, to keep N3 running to have enough good tiles for Arrow Lake mass production ramp up.

 

[Saylick]

WCCFtech has the HC34 Intel Meteor Lake slide deck. Warning: Image dump below:

 

[Saylick]

 

[Saylick]

 

[Saylick]

 

[igor_kavinski]

Are they talking about Meteor Lake here? Or is the architecture the same for MTL (mobile) and Arrow Lake (desktop)?

 

[inf64]

View attachment 66462

Are they talking about Meteor Lake here? Or is the architecture the same for MTL (mobile) and Arrow Lake (desktop)?

The CPU tile is red for Arrow Lake so I assume the core is next-gen (successor of Meteor Lake).

 

[IntelUser2000]

View attachment 66462

Are they talking about Meteor Lake here? Or is the architecture the same for MTL (mobile) and Arrow Lake (desktop)?

I don't think it should be taken as anything specific. They are just saying it allows them to be that flexible.

@Exist50 The number of Xe cores being 4-16 shown above equals 64-192 EUs. If we believe some leaks are intentional we can also speculate it's meant to obfuscate.

The slides confirm Foveros Omni starts at 25um. It even suggests Lunar Lake might be Foveros, not Omni

 

[Joe NYC]

There doesn't seem to be anything wrong with it now. It's just very overpriced. That was the strange part about using N3 - a $40 IGP chiplet makes zero sense.

The one thing I do wonder about is that if they are really using 22FFL for the base tile, where they would get the capacity. I imagine they probably still have the tools but they would need a lot of space.

It does seem like an overkill for iGPU.

Another thing it tells me that the days of nVidia selling notebook dGPUs are likely numbered.

 

[mikk]

Core counts scalability tops out at 6+8, I think this will be the max core count on desktop. It makes sense considering the rumor says MTL is only for lowend-midrange on desktop. GPU scalability goes from 64EUs up to 192 EUs.

 

[Saylick]

It does seem like an overkill for iGPU.

Another thing it tells me that the days of nVidia selling notebook dGPUs are likely numbered.

Nvidia won't be completely ousted from the notebook dGPU market. For CAD workstations, they'll maintain their lock due to excellent drivers + Optix. They might lose the lower end of the notebook gaming market, but it will be much harder to lose the higher end.

 

[witeken]

Intel, America's biggest chipmaker, is working with TSMC on at least two 3-nm projects to design central processing units for notebooks and data center servers in an attempt to regain market share it has lost to Advanced Micro Devices and Nvidia over the past few years. Mass production of these chips is expected to begin by the end of 2022 at the earliest.
"Currently the chip volume planned for Intel is more than that for Apple's iPad using the 3-nanometer process," one of the sources said.

Apple and Intel become first to adopt TSMC's latest chip tech

Taiwanese company remains vital partner for American tech giants

asia.nikkei.com

Psst: Reread your links as Arrow Lake. Those links do not mention Meteor Lake (although, the 2021 link doesn't work, so I didn't read it).

In order to have Arrow Lake in late 2023/early 2024, Intel needs N3 up and running in 2022 for engineering and qualification samples. Then, once they pass, to keep N3 running to have enough good tiles for Arrow Lake mass production ramp up.

You've been warned before yet i'll inform you once again. You must include your own personal commentary when dropping links or other members quotes.

Daveybrat
AT Moderator

 

[igor_kavinski]

Mass production of these chips is expected to begin by the end of 2022 at the earliest.

If this is true, laptop chiplets built with N3 could be out by end of June 2023. So tGPU will be on N3?

 

[burninatortech4]

Ultra low voltage core. In the past that was a way to refer to the U series chips. Though in this case it may refer to a (new?) core in the SoC used as the system controller.

So, similar to the Quark core on the Lewisburg PCH?

 

[Exist50]

If this is true, laptop chiplets built with N3 could be out by end of June 2023. So tGPU will be on N3?

Nothing about Meteor Lake uses N3.

 

[nicalandia]

Have we seen any indication yet that Intel is preparing an 8+16 Meteor Lake CPU tile?

 

[igor_kavinski]

Nothing about Meteor Lake uses N3.

What is Intel making on N3 for mobile?