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.)

dttprofessor · Tuesday at 9:55 AM

511 said:
Everything is finalized for NVL at this point there are only 4 tiles
8+16/8+16 bLLC
4+8/4+0

LPE back to soc tile ?

511 · Tuesday at 10:01 AM

dttprofessor said:
LPE back to soc tile ?

Yes this is already leaked by leaker somewhere the SoC contains LP-E/IMC, Xe4 Display/Media,
Die waste Unit

dttprofessor · Tuesday at 10:08 AM

511 said:
Yes this is already leaked by leaker somewhere the SoC contains LP-E/IMC

Which node they use for NVL soc tile ?
3?18a?
Which connection for the separated IMC ？

511 · Tuesday at 10:21 AM

dttprofessor said:
Which node they use for NVL soc tile ?
3?18a?
Which connection for the separated IMC ？

Don't know

dttprofessor · Tuesday at 10:25 AM

511 said:
Don't know

The soc tile of NVLs is the type of new n100?

LightningZ71 · Tuesday at 11:06 AM

My post was more conjectural. They could do a stack with only 4 tiles if they wanted. 12P + slightly enhanced L3, 40E, 6+8, 2+8. Low performance tiles don't need to be optimized, and could be on trailing nodes with optimized libraries instead to lower costs as people in those segments don't care. If I really needed all those threads, I'd rather 12 optimized P cores and 40 optimized E cores instead of of 16p/32e that have to suffer compromise. Plus, now all the P cores live in the same L3 cache domain, and anything that's so parallel that it needs a sea of E cores will have no problem that all the E cores are on their own die anyway. Best of both worlds.

Again, this is a "could have been"

511 · Tuesday at 11:39 AM

It requires more design resources than simple 2*(8+16)

Josh128 · Tuesday at 12:42 PM

511 said:
Yes this is already leaked by leaker somewhere the SoC contains LP-E/IMC, Xe4 Display/Media,
Die waste Unit

Die waste unit?

511 · Tuesday at 1:04 PM

Josh128 said:
Die waste unit?

NPU

igor_kavinski · Tuesday at 1:38 PM

511 said:
intel should release a handheld with 8 Darkmont and 10-12 Xe3 Cores.

They rarely do what they should. There is still no successor to the Core i3-N305 laptops. It's like Skylake is a flippin' curse that won't end unless the computing experience of millions of souls is ruined first.

ondma · Tuesday at 2:25 PM

511 said:
It requires more design resources than simple 2*(8+16)

Havent been following CPUs much lately, due to Intel not being very competitive. What is the consensus: Is NL going to really double the core counts, or is this just another rumor/hope that will ultimately get cancelled like so many other projects from Intel lately? It sounds great, but upon thinking about it, what would this accomplish? How would they manage the heat and power usage, unless the new core design is hugely more efficient. I know they will be on a more efficient process, but that couldnt possibly save enough power to double the core counts. The problem with ARL seems to be latency, not raw processing power anyway. If they do this, and can get the power and temps under control, it would obviously be a multi threading beast, but for gaming??

511 · Tuesday at 2:34 PM

igor_kavinski said:
They rarely do what they should. There is still no successor to the Core i3-N305 laptops. It's like Skylake is a flippin' curse that won't end unless the computing experience of millions of souls is ruined first.

Wild Cat Lake exists and would launch next year

ondma said:
Havent been following CPUs much lately, due to Intel not being very competitive. What is the consensus: Is NL going to really double the core counts, or is this just another rumor/hope that will ultimately get cancelled like so many other projects from Intel lately? It sounds great, but upon thinking about it, what would this accomplish? How would they manage the heat and power usage, unless the new core design is hugely more efficient. I know they will be on a more efficient process, but that couldnt possibly save enough power to double the core counts. The problem with ARL seems to be latency, not raw processing power anyway. If they do this, and can get the power and temps under control, it would obviously be a multi threading beast, but for gaming??

Yeah it's happening 2X the core count vs ARL not to mention AVX 10.2 with 512 bit vector aka AVX-512.

reb0rn · Tuesday at 3:42 PM

As someone who need more core also ppl that have AMD, I am quite eager to learn perf/watt of the new CPU, I am ready to sell all ZEN 4/5 CPU and move on to intel in like a 10y if its better in that regard by 20%+ (preferable perf.watt at least 25% better then zen in multicore in limited TDP which give you best ratio and ok perf as now I have zen 5 at 140w limit)
I know no one has numbers and most ppl do not believe after few not se gr8 nodes upgrades its possible, even leaks are none where we see multicore perf per watt at all

jpiniero · Tuesday at 3:49 PM

511 said:
NPU

Stonks tho

Joe NYC · Tuesday at 4:40 PM

511 said:
Everything is finalized for NVL at this point there are only 4 tiles
8+16/8+16 bLLC
4+8/4+0

Thanks, that's helpful.

So that's already 4 dies, and there is no P-core only and E-Core only tile...

I don't know if I understood Raichu correctly, but he seemed to be saying that Intel may sell SKU with only bLLC tile, which could be an interesting product.

Josh128 · Tuesday at 5:01 PM

reb0rn said:
As someone who need more core also ppl that have AMD, I am quite eager to learn perf/watt of the new CPU, I am ready to sell all ZEN 4/5 CPU and move on to intel in like a 10y if its better in that regard by 20%+ (preferable perf.watt at least 25% better then zen in multicore in limited TDP which give you best ratio and ok perf as now I have zen 5 at 140w limit)
I know no one has numbers and most ppl do not believe after few not se gr8 nodes upgrades its possible, even leaks are none where we see multicore perf per watt at all

10 years?! Bro, its possible none of us will even be here in 10 years.

igor_kavinski · Tuesday at 5:04 PM

Josh128 said:
10 years?! Bro, its possible none of us will even be here in 10 years.

More importantly, is Intel gonna be here in 10 years? Is x86 gonna be a thing in 10 years? Very clouded the future is.

reb0rn · Tuesday at 7:16 PM

Josh128 said:
10 years?! Bro, its possible none of us will even be here in 10 years.

No I mean I am almost 10y on AMD its less 8Y I think, its time we shift back at least for multicore and I but it gona happen

Josh128 · Tuesday at 9:24 PM

igor_kavinski said:
More importantly, is Intel gonna be here in 10 years? Is x86 gonna be a thing in 10 years? Very clouded the future is.

511 · Tuesday at 10:51 PM

igor_kavinski said:
More importantly, is Intel gonna be here in 10 years? Is x86 gonna be a thing in 10 years? Very clouded the future is.

US government is gonna make sure Intel foundry will exist lol

igor_kavinski · Tuesday at 10:58 PM

511 said:
US government is gonna make sure Intel foundry will exist lol

Or Trump could sell it to the Chinese in return for no tariffs on them.

"You take the foundries for $100 billion and you promise to invest $10 billion every year in the expansion/upkeep of the foundries and I won't tariff you guys".

Then he goes, look! I made the biggest deal of the century. I'm so smart!

Thibsie · Tuesday at 11:32 PM

511 said:
US government is gonna make sure Intel foundry will exist lol

I fear USG has no clue at all, either about foundry or economics btw.

vanplayer · Tuesday at 11:41 PM

ondma said:
How would they manage the heat and power usage, unless the new core design is hugely more efficient.

It's hard to imagine typical heatsink solution can cool this thing. This is the only thing I know currently.

Well, I don't know whether those leakers are tight-lipped or not, maybe even they weren't so sure, but I guess this thing either suffers various technical issues that Intel decided to cover it up at all cost, or it doesn't tape out yet, it's still far far away from release, maybe 1H27 if it doen't accelerate.

coercitiv · Wednesday at 2:09 AM

ondma said:
Havent been following CPUs much lately, due to Intel not being very competitive. What is the consensus: Is NL going to really double the core counts, or is this just another rumor/hope that will ultimately get cancelled like so many other projects from Intel lately? It sounds great, but upon thinking about it, what would this accomplish? How would they manage the heat and power usage, unless the new core design is hugely more efficient. I know they will be on a more efficient process, but that couldnt possibly save enough power to double the core counts. The problem with ARL seems to be latency, not raw processing power anyway. If they do this, and can get the power and temps under control, it would obviously be a multi threading beast, but for gaming??

I'll go quickly through your questions:

Is NL going to really double the core counts - Very unlikely to happen to their whole lineup, possible to happen for one/few ultra premium SKUs (ultra pun intended). Some of this core count increase may trickle down to value SKUs, but only a fraction.
Is this just another rumor/hope that will ultimately get cancelled - I think canceling it is in the books, depends a lot on Intel's health and internal estimates of what they've accomplished.
What would this accomplish - Two things: halo product that brings some much needed attention (especially if it features a bigger LLC to help hide latency), test & learning vehicle for future products where they break the compute tile further. Even with the current tile setup, Intel's compute tile is huge in compared to the competition (~115mm2 vs 71mm2), they're probably exploring ways to lower cost and be more agile.
How would they manage the heat and power usage - heat and power are the least of their problems. Even Arrow Lake in a 2 x 8+16 configuration would do great, as each compute die would use ~120W and operate in a good efficiency spot. Think about it this way, every Intel desktop CPU since Alder Lake had much higher efficiency @120W than @240W, so doubling the cores would be a good use of the socket power limit. Temperature would also be perfectly fine, cores would operate at lower voltages and heat would be spread on 2 compute dies instead of one (think around 2x100+mm2 for just the compute dies). Intel can always mess this up with another 300W+ "enthusiast" profile though.
The problem with ARL seems to be latency, not raw processing power anyway - Yes, they need more time to adapt.
It would obviously be a multi threading beast, but for gaming?? - If they use an alternative to V-cache, the product will be better for gaming than ARL. It would still be an awkward setup, because the P cores will be spread on 2 dies, so gaming performance will be optimal only when parking one compute die.

The biggest problem we highlighted a few pages back is cost. This product will be expensive to make and hard to sell for a huge premium. It also requires good sync and execution from multiple Intel teams to pull off just for the sake of a halo product. It may not make the cut based on Intel's new 50% gross margin policy. It would certainly be fun and exciting for us though, so keep calm and fingers crossed.

ondma · Wednesday at 2:38 AM

coercitiv said:
I'll go quickly through your questions:

Is NL going to really double the core counts - Very unlikely to happen to their whole lineup, possible to happen for one/few ultra premium SKUs (ultra pun intended). Some of this core count increase may trickle down to value SKUs, but only a fraction.

Is this just another rumor/hope that will ultimately get cancelled - I think canceling it is in the books, depends a lot on Intel's health and internal estimates of what they've accomplished.

What would this accomplish - Two things: halo product that brings some much needed attention (especially if it features a bigger LLC to help hide latency), test & learning vehicle for future products where they break the compute tile further. Even with the current tile setup, Intel's compute tile is huge in compared to the competition (~115mm2 vs 71mm2), they're probably exploring ways to lower cost and be more agile.

How would they manage the heat and power usage - heat and power are the least of their problems. Even Arrow Lake in a 2 x 8+16 configuration would do great, as each compute die would use ~120W and operate in a good efficiency spot. Think about it this way, every Intel desktop CPU since Alder Lake had much higher efficiency @120W than @240W, so doubling the cores would be a good use of the socket power limit. Temperature would also be perfectly fine, cores would operate at lower voltages and heat would be spread on 2 compute dies instead of one (think around 2x100+mm2 for just the compute dies). Intel can always mess this up with another 300W+ "enthusiast" profile though.

The problem with ARL seems to be latency, not raw processing power anyway - Yes, they need more time to adapt.

It would obviously be a multi threading beast, but for gaming?? - If they use an alternative to V-cache, the product will be better for gaming than ARL. It would still be an awkward setup, because the P cores will be spread on 2 dies, so gaming performance will be optimal only when parking one compute die.

The biggest problem we highlighted a few pages back is cost. This product will be expensive to make and hard to sell for a huge premium. It also requires good sync and execution from multiple Intel teams to pull off just for the sake of a halo product. It may not make the cut based on Intel's new 50% gross margin policy. It would certainly be fun and exciting for us though, so keep calm and fingers crossed.

Thanks for the reply. As far as heat and power, seems like your point 4 would lead to lower clock speeds and possibly a sacrifice in gaming performance, which is my primary interest. I really dont think Intel can catch up in gaming until they come up with something analogous to v-cache. (Or make a huge Conroe like jump in P core performance, which seems very unlikely.)

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

Senior member

Attachments

Member

Platinum Member

Member

Platinum Member

Member

Platinum Member

Platinum Member

Senior member

Platinum Member

Lifer

Diamond Member

Platinum Member

Senior member

Lifer

Diamond Member

Senior member

Lifer

Senior member

Senior member

Platinum Member

Lifer

Golden Member

Member

Diamond Member

Diamond Member