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.





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.)

 

[controlflow]

Lunar lake took approx 1 year from power on to availability 2 Steppings
Meteor lake took 1.5-2 years with 3 steppings
Upto you which data point to take
Both of them were with Node Delays N3B was 6 month delayed and Intel 7nm(Intel 4) was delayed for a couple of years

BTW How did TSMC managed to get 6 month delay due to Intel ?🤣🤣

This isn't the actual "power on". Power on for for CWF and PTL happened before August. They were booting OS months ago: https://www.tomshardware.com/pc-com...r-lake-and-clearwater-forest-cpus-are-booting

The recent reports are referring to ES PTL samples at OEMs being used.

 

[511]

This isn't the actual "power on". Power on for for CWF and PTL happened before August. They were booting OS months ago: https://www.tomshardware.com/pc-com...r-lake-and-clearwater-forest-cpus-are-booting

The recent reports are referring to ES PTL samples at OEMs being used.

Yes but i am referring to intel powering on at Intel Labs

 

[511]

Intel "Alder Lake" the new "Skylake" and "Intel 10nm" the new "Intel 14nm"

https://twitter.com/x/status/1867953437192929652

Editing+
Editing++

 

[igor_kavinski]

Linux 6.13-rc3 fixes expensive CPUID handling on Sapphire Rapids CPUs — an issue resulting in Skylake CPUs being 4X faster for CPUID-related updates

This is arguably one of the most bizarre issues to crop up to date in the Linux kernel.

www.tomshardware.com

Apparently all those Sapphire Rapids steppings and this issue still managed to creep in. First of all, it should never have happened if they had copy pasted the relevant silicon blocks from tried and true previous designs. Second, looks like they are using crappy validation teams. So nothing has changed since Skylake.

 

[SiliconFly]

.

 

[KompuKare]

The magic firmware upgrade will fix ARL - especially in games - always has an extreme copium vibe to it IMO.

About the only way I might have bought it, is it we had tons of rumours of Intel launching ARL nearly totally unready. We heard some rumours of a slightly rushed launch but not enough to justify the copium.

 

[Hulk]

An early leak on the new µcode update (0x114) for Intel Arrow Lake desktop CPUs. And it looks disappointing.

https://twitter.com/x/status/1869095694348562653

Sadly, it appears ARL-S desktop parts are a dud after all. Hope at least the upcoming ARL-H laptop parts are able to stand up to competition as they don't need to be clocked high.

My prediction for the Intel response to microcode update testing.

"Unfortunately the results obtained by indpendent reviewers of ARL with the microcode update were not what we expected."

Oh, okay. That explains everything. Thanks Intel. Really, thanks for continuing to be so forthcoming. Your transparency is quite a relief to all of the customers who have purchased your processors.

 

[SiliconFly]

.

 

[SiliconFly]

.

 

[Thunder 57]

An early leak on the new µcode update (0x114) for Intel Arrow Lake desktop CPUs. And it looks disappointing.

https://twitter.com/x/status/1869095694348562653

Sadly, it appears ARL-S desktop parts are a dud after all. Hope at least the upcoming ARL-H laptop parts are able to stand up to competition as they don't need to be clocked high.

In my experience updates, whether it be microcode, BIOS, Windows, etc have never amounted to much of anything. You just seem to hear that whenever a lackluster part is released. It reminds me of the Bulldozer days when they said an update for the Windows scheduler was needed. It did next to nothing.

 

[511]

Linux 6.13-rc3 fixes expensive CPUID handling on Sapphire Rapids CPUs — an issue resulting in Skylake CPUs being 4X faster for CPUID-related updates

This is arguably one of the most bizarre issues to crop up to date in the Linux kernel.

www.tomshardware.com

Apparently all those Sapphire Rapids steppings and this issue still managed to creep in. First of all, it should never have happened if they had copy pasted the relevant silicon blocks from tried and true previous designs. Second, looks like they are using crappy validation teams. So nothing has changed since Skylake.

EMR has less stepping than SPR LOL
It's 2 vs 8! Steppings
GNR is only 1 stepping

 

[DavidC1]

Apparently all those Sapphire Rapids steppings and this issue still managed to creep in. First of all, it should never have happened if they had copy pasted the relevant silicon blocks from tried and true previous designs. Second, looks like they are using crappy validation teams. So nothing has changed since Skylake.

It's not crappy validation teams. At one point they canned the entire team. That's why it got delayed, and why it struggled.

Also on a device as complex as a modern high end CPU core, there's no such thing as pure copy and paste. It's not Windows Paint.

 

[SiliconFly]

.

 

[Hulk]

EMR has less stepping than SPR LOL
It's 2 vs 8! Steppings
GNR is only 1 stepping
View attachment 113394View attachment 113395

Too small, can't read it.

 

[Hulk]

Turns out this was premature exaggeration.

Intel's Robert Hallock is about to go live in 30 mins to give us a fresh update on the still upcoming ARL bios+µcode update/performance fix (probably 0x115). Should be interesting. Stay tuned.

Time to start dancing.

 

[511]

Too small, can't read it.

Too small 🤣

Anyways here is a readable version

1 or 2 stepping is industry standard

 

[511]

New Bios in January what is with the cope

 

[Hulk]

 

[SiliconFly]

.

 

[511]

View attachment 113428

Nooooo i was going to post it anyway i will post this first in January

 

[Saylick]

View attachment 113428

Jesus... What a sh*tshow.

First it was firmware/BIOS updates to fix Raptor Lake degradation and now this.

 

[Kepler_L2]

View attachment 113428

Up to 58% performance increase with Windows/BIOS updates?

 

[511]

Up to 58% performance increase with Windows/BIOS updates?

AMD we identified a bug fixing this would improve performance by 5%
Intel: Hold my Bear we had BIOS and SW optimization we can improve performance by upto 58%

 

[SiliconFly]

.

 

[511]

They said single digit so calm your horses lol