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.)

alcoholbob · May 9, 2025

I can't imagine NVL latency is going to be any good with a dual CCD problems...if anything I imagine it's going to be as bad for gaming as Arrow Lake and Raptor Lake remains the gaming champ for Intel well into 2027-2028. 52 cores screams going all-in on all-core synthetic benchmarks and losing at everything else.

Thunder 57 · May 10, 2025

Fjodor2001 said:
As I wrote, not in all workloads. And different workloads will benefit more from it than others. But over time, more applications will be adapted to make better use of more cores, the more cores are commonly available.

We’ve already seen this trend over a long time. Compare the applications and games from 10+ years to what we have today and see how much they have improved in this regard already.

There”s really not many other options than to increase performance via increased core count currently, since we’re not getting much perf increase from IPC or frequency increase anymore.

As has been said, Amdahl's Law.

511 · May 10, 2025

alcoholbob said:
I can't imagine NVL latency is going to be any good with a dual CCD problems...if anything I imagine it's going to be as bad for gaming as Arrow Lake and Raptor Lake remains the gaming champ for Intel well into 2027-2028. 52 cores screams going all-in on all-core synthetic benchmarks and losing at everything else.

They just need to re design the fabric

alcoholbob · May 10, 2025

511 said:
They just need to re design the fabric

And DDR5-12000 CUDIMMS probably...just to get close to Raptor Lake

511 · May 10, 2025

alcoholbob said:
And DDR5-12000 CUDIMMS probably...just to get close to Raptor Lake

bruh like i have said before the packing they use is similar to Hopper so packing is not the bottleneck the fabric is and L3 as well i don't think Intel's Designers are so dumb and as fir DDR5 CUDIMMS at 12K that's going to be OC only.

msj10 · May 10, 2025

You just need to look at Lunar Lake to get an idea about NVL memory latency. It's not that hard to figure out.

511 · May 10, 2025

msj10 said:
You just need to look at Lunar Lake to get an idea about NVL memory latency. It's not that hard to figure out.

LNL is monolithic lol the PCH is combined using foveros that's it PTL you can call chiplet

msj10 · May 10, 2025

511 said:
LNL is monolithic lol the PCH is combined using foveros that's it PTL you can call chiplet

NVL will be based on the LNL soc architecture. Intel even said the architecture can support multiple compute dies in the coming generations. they were obviously referring to NVL.

511 · May 10, 2025

I have high hopes if Glenn Hilton is working on it.

OneEng2 · May 10, 2025

511 said:
They just need to re design the fabric

I think it is likely a bit more complex than that. I think the cores themselves need to be architected in a way that reduces the impact of off-chip latency on the processor performance. I do agree with your assessment that it is highly likely that the fabric will also be improved to lower the actual latency as well. This is easily the lowest hanging fruit as the core re-architecture will certainly be harder to achieve.

IEC · May 10, 2025

Lunar Lake is quite nice for a portable laptop with good battery life.

This is my 226V in an ASUS Vivobook 14S:

511 · May 10, 2025

Lunar Lake is one of the best Product Intel has launched in recent years.

511 · May 11, 2025

Btw does there exist a copilot hate thread cause what I am hearing is MSFT being ridiculous.
Requiring 80TOPS NPU

511 · May 11, 2025

https://twitter.com/x/status/1921525230482145439

Fjodor2001 · May 12, 2025

Thunder 57 said:
As has been said, Amdahl's Law.

As has been said, exactly the same thing that was mentioned by some during the 4C-forever phase. We passed that, and we will pass the 16C-forever phase too.

Tigerick · May 12, 2025

511 said:
Btw does there exist a copilot hate thread cause what I am hearing is MSFT being ridiculous.
Requiring 80TOPS NPU

Source ?

inquiss · May 12, 2025

Fjodor2001 said:
As has been said, exactly the same thing that was mentioned by some during the 4C-forever phase. We passed that, and we will pass the 16C-forever phase too.

We're barely in the 16c phase. Most people want 8c or less. And the point of the law is you get diminishing benefits for the increased number of cores, so the position is different from 4 to 8 and then 8 to 16 and less benefit going up after that.

I'm not saying you're wrong over the long run, but honestly consumer doesn't need more cores. They need faster cores and bigger caches.

Fjodor2001 · May 12, 2025

inquiss said:
We're barely in the 16c phase. Most people want 8c or less. And the point of the law is you get diminishing benefits for the increased number of cores, so the position is different from 4 to 8 and then 8 to 16 and less benefit going up after that.

I'm not saying you're wrong over the long run, but honestly consumer doesn't need more cores. They need faster cores and bigger caches.

Consumer CPUs have maxed out at 16C (or 24C for Intel) for several years, same as when we were stuck at 4C for several years.

In the 4C-forever era it was said by some that nobody needed or wanted more than 4C, and it would be pointless due to Amdahl’s law, so no point providing more than that to the consumer. Then AMD proved that to be wrong, and introduced 8C and soon after 16C CPUs. They did so to break the Intel lead at that time.

This time it’s AMD that is in the lead, but stubbornly stuck at 16C. The exact same arguments are mentioned by the ST perf is all that matters crowd. And I think this time the tables will turn so it’ll be Intel that bumps core count to prove them wrong. Rumors say it’ll be to 52C, although with a mix of P/E/LPE cores.

There really is not many other options to substantialy increase perf than via core count increase, since IPC and frequency increase are only improving perf by a measly ~5-7% per CPU generation currently.

MS_AT · May 12, 2025

Fjodor2001 said:
As has been said, exactly the same thing that was mentioned by some during the 4C-forever phase. We passed that, and we will pass the 16C-forever phase too.

Sure, get proper GPU slicing and host game streaming for your whole family to their mobile devices. This will scale plenty with cores, and if you run into scaling limit, make the family bigger. This way we can scale quite well, depending of course on how many relatives you are willing to deal with, and it might be the house starts to become limitation faster than the cores, but well, one problem at at time

inquiss · May 12, 2025

Fjodor2001 said:
Consumer CPUs have maxed out at 16C (or 24C for Intel) for several years, same as when we were stuck at 4C for several years.

In the 4C-forever era it was said by some that nobody needed or wanted more than 4C, and it would be pointless due to Amdahl’s law, so no point providing more than that to the consumer. Then AMD proved that to be wrong, and introduced 8C and soon after 16C CPUs. They did so to break the Intel lead at that time.

This time it’s AMD that is in the lead, but stubbornly stuck at 16C. The exact same arguments are mentioned by the ST perf is all that matters crowd. And I think this time the tables will turn so it’ll be Intel that bumps core count to prove them wrong. Rumors say it’ll be to 52C, although with a mix of P/E/LPE cores.

How many people buy 16c products. The market is small compared to mainstream.

On your second point, you get more from going from 4 to 8 than 8 to 16. There's mainly for people with embarrassingly parallel workloads, not common workloads. What's the more common core count sold today? It's not 16. It's been 6, and maybe recently 8 and only because the x3D chips have 8.

Intel could release the rumoured 52c product. Doesn't mean it will be popular. It won't be. I expect that to be cancelled, though your mileage might vary. If you only run cinebench it might be good at that.

The x3D chips show that cache and speed matter more. You'll get more than 16 cores from AMD in the next platform..it's convenient to do so when reusing server designs. But bumping it up massively for client? Massive meh

Fjodor2001 · May 12, 2025

MS_AT said:
Sure, get proper GPU slicing and host game streaming for your whole family to their mobile devices. This will scale plenty with cores, and if you run into scaling limit, make the family bigger. This way we can scale quite well, depending of course on how many relatives you are willing to deal with, and it might be the house starts to become limitation faster than the cores, but well, one problem at at time

Again, same aguments as during the 4C-forever area, when it was said that there was no point in going beyond 4C.

reb0rn · May 12, 2025

As soon 16 core is cheap and mainstream price they will buy it, while some % will want 32-64 core parts........ its just most ppl just can not see the reality of being stuck in time

Fjodor2001 · May 12, 2025

inquiss said:
How many people buy 16c products. The market is small compared to mainstream.

On your second point, you get more from going from 4 to 8 than 8 to 16. There's mainly for people with embarrassingly parallel workloads, not common workloads. What's the more common core count sold today? It's not 16. It's been 6, and maybe recently 8 and only because the x3D chips have 8.

Intel could release the rumoured 52c product. Doesn't mean it will be popular. It won't be. I expect that to be cancelled, though your mileage might vary. If you only run cinebench it might be good at that.

The x3D chips show that cache and speed matter more. You'll get more than 16 cores from AMD in the next platform..it's convenient to do so when reusing server designs. But bumping it up massively for client? Massive meh

How do you intend to improve perf substantially if not via core count increase?

Are you satisfied with 5-7% perf increase via IPC/frequency per CPU generation?

inquiss · May 12, 2025

Fjodor2001 said:
How do you intend to improve perf substantially if not via core count increase?

Are you satisfied with 5-7% perf increase via IPC/frequency per CPU generation?

Architecture, clock speed, cache size and cache structure, packaging, new instructions, IPC. Loads of options.

No, but I don't get why you think core counts are the only way to achieve something greater than 5-7% per increase per generation.

Hitman928 · May 12, 2025

Fjodor2001 said:
Again, same aguments as during the 4C-forever area, when it was said that there was no point in going beyond 4C.

I don't remember many people making this argument at the time outside of maybe some Intel supporters who didn't want any criticism against their favorite company. That's not the situation today. Additionally, as soon as a viable 6+ core option was given, it sold extremely well. We have 16+ core SKUs today that don't sell all that well relative to their lower core contemporaries and in many instances they aren't sold because they have 16+ cores but because they have the best binned cores. If there was really demand for really high core count CPUs, we'd see it in the sales figures pushing the currently highest core SKUs up the sales charts. That's not to say there isn't some demand and every user is welcome to advocate for their own wants, but don't confuse niche wants for what the market is actually asking for.

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

Senior member

Attachments

Diamond Member

Diamond Member

Platinum Member

Diamond Member

Platinum Member

Member

Platinum Member

Member

Platinum Member

Senior member

Elite Member

Platinum Member

Platinum Member

Platinum Member

Diamond Member

Senior member

Senior member

Diamond Member

Senior member

Senior member

Diamond Member

Senior member

Diamond Member

Senior member

Diamond Member