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

Page 858 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Tigerick

Senior member
Apr 1, 2022
816
781
106
PPT1.jpg
PPT2.jpg
PPT3.jpg



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.



LNL-MX.png

Intel Core Ultra 100 - Meteor Lake

INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg

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.)



Clockspeed.png
 

Attachments

  • PantherLake.png
    PantherLake.png
    283.5 KB · Views: 24,026
  • LNL.png
    LNL.png
    881.8 KB · Views: 25,519
Last edited:

poke01

Diamond Member
Mar 8, 2022
4,036
5,362
106

ToTTenTranz

Senior member
Feb 4, 2021
578
1,033
136
Interesting they don’t mention “faster cpu”.
They're clear about a faster GPU, though.

I do wonder how Panther Lake's Xe3 will behave against Strix Point's 16CU RDNA3.5, as the 140V and 140T are already pretty competitive with that model.

It does look like Intel will have an iGPU performance crown on the 15-35W range, at least until Medusa Premium comes out.
 

511

Diamond Member
Jul 12, 2024
3,915
3,656
106
They're clear about a faster GPU, though.

I do wonder how Panther Lake's Xe3 will behave against Strix Point's 16CU RDNA3.5, as the 140V and 140T are already pretty competitive with that model.

It does look like Intel will have an iGPU performance crown on the 15-35W range, at least until Medusa Premium comes out.
Medusa Premium has how much CU and which arch? also that would likely compete with NVL iGPU.
 

ToTTenTranz

Senior member
Feb 4, 2021
578
1,033
136
Medusa Premium has how much CU and which arch?
Medusa Premium has an AT4, so 12 RDNA5 CUs (supposedly close to 24 RDNA4 CUs in throughput per clock).

also that would likely compete with NVL iGPU.

I think we only know the specs for N1X so far, which is way too big and will probably be closer to Medusa Halo. There's also a "N1" for laptops and perhaps that one will be competition to Medusa Premium.
 
  • Like
Reactions: Tlh97 and 511

511

Diamond Member
Jul 12, 2024
3,915
3,656
106
  • Like
Reactions: DKR

DavidC1

Golden Member
Dec 29, 2023
1,792
2,893
96
If what that reddit guy is saying is true and that Darkmont is quite a bit more efficient ISO-process(~10%) and 18A is better, then it can potentially counterract losses from Lunarlake from a battery life point of view. In Lunarlake, Skymont LPE is half of the reason it gets good battery life. Pantherlake may still lose in light load scenarios, but if Darkmont makes it up in less lighter loads such as Teams then overall they can call it similar.

I know that even for Alderlake laptops you can get great battery life, but you need to go through lot of work to get there. Cut all the services running to a minimum, have Task Manager up frequently to check for mis-behaving applications and find a way to terminate them or make them work properly, have all basic power management settings work such as Drive power savings, plus get drivers of every single component in your laptop updated to the latest. In reality, that is not feasible as vanishingly small amount of people know how to do this. I don't even have more than basic MS antivirus enabled for this very reason, because I know it's much harder to get 3rd party antivirus software working with low demand enough to keep the CPU utilization under 3% for the application.

The other day I was frustrated with my 10th Gen Pentium Desktop struggling with OpenOffice, when computers of 40 years ago could do the same. Some articles were talking about increasing abstraction as a culprit. The whole computer market is a pseudo-ponzi scheme.

Having a dedicated power efficient core offload things so you don't need to to the above improves things for the average scenario, because inevitably the demand rises due to inefficiency in software or whatever other reasons.
 
  • Haha
Reactions: 511
Jul 27, 2020
27,225
18,718
146

Geddagod

Golden Member
Dec 28, 2021
1,513
1,611
106
Just a thought, one has to speculate about Intel shortening the pipeline of their P-cores as a pretty good thing. Keeping the same high Fmax, while also having the side effects of slight IPC increases, and lowering the need to spend a bunch of resources on the BPU, since the branch mispredict penalty will also be lower.
 

511

Diamond Member
Jul 12, 2024
3,915
3,656
106
Just a thought, one has to speculate about Intel shortening the pipeline of their P-cores as a pretty good thing. Keeping the same high Fmax, while also having the side effects of slight IPC increases, and lowering the need to spend a bunch of resources on the BPU, since the branch mispredict penalty will also be lower.
PNC is clustered Decode with shared L2 it's more like Atom in Nova than a P core
 

Magio

Member
May 13, 2024
155
176
76

GB 6.5 is suspiciously close to Panther Lake launch. Could this mean Panther Lake supports APX?

GB 6.5 is equally close to A19 Pro and 8 Elite Gen 2, tho (who knows, maybe they'll finally rebalance the benchmark now that non-Apple chips will benefit from the SME boost). I think up to now rumors were that APX would only be in Nova Lake?
 
  • Like
Reactions: igor_kavinski

511

Diamond Member
Jul 12, 2024
3,915
3,656
106
Dude!

Both FRED and LKGS deal with minimizing the context switch latency so Intel could gain substantial wins in multicore I/O heavy workloads (think streaming games especially!).
This needs to supported by OS as well knowing windows see you in couple of years on Linux we should be fine.