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.)

 

[Doug S]

Let's see. M4 does not run Linux or Windows.

M4 can run Windows - but so far only inside a VM since Microsoft's exclusivity agreement with Qualcomm prevents them from booting it even if they wanted to. Though after that unless someone writes a Windows GPU driver it'll never boot natively, but servers run headless so the lack of a GPU driver wouldn't matter for a hypothetical Apple Server.

As for Linux, I guess you aren't aware that Apple Silicon Macs ARE supported by Asahi Linux. That support isn't perfect, but the shortcomings are things that don't matter for servers like full native GPU performance, microphone issues and so forth.

If Apple was going to enter the server world they could deliver the necessary drivers for Linux and Windows Server without much trouble. You're pretending this would be some sort of herculean effort. Its trivial compared to what they accomplished when they migrated the entire Mac world from x86 to ARM.

 

[Io Magnesso]

M4 can run Windows - but so far only inside a VM since Microsoft's exclusivity agreement with Qualcomm prevents them from booting it even if they wanted to. Though after that unless someone writes a Windows GPU driver it'll never boot natively, but servers run headless so the lack of a GPU driver wouldn't matter for a hypothetical Apple Server.

As for Linux, I guess you aren't aware that Apple Silicon Macs ARE supported by Asahi Linux. That support isn't perfect, but the shortcomings are things that don't matter for servers like full native GPU performance, microphone issues and so forth.

If Apple was going to enter the server world they could deliver the necessary drivers for Linux and Windows Server without much trouble. You're pretending this would be some sort of herculean effort. Its trivial compared to what they accomplished when they migrated the entire Mac world from x86 to ARM.

Honestly, Asahi Linux I think it's better to know the hardships of the creators about...
Only the M1 series released 5 years ago is completely moving now.
The M2 series is currently working on it, but it will still take some time.
M3/M4 takes more time...

 

[Io Magnesso]

Honestly, Asahi Linux I think it's better to know the hardships of the creators about...
Only the M1 series released 5 years ago is completely moving now.
The M2 series is currently working on it, but it will still take some time.
M3/M4 takes more time...

Since there is no information, it is being developed by reverse engineering from 1.
It's really hard...
I don't think Apple, which freerides to other ecosystems or platforms, I don't think I can lend a hand
Well, the story of Linux and Windows is for example...

 

[poke01]

Honestly, Asahi Linux I think it's better to know the hardships of the creators about...
Only the M1 series released 5 years ago is completely moving now.
The M2 series is currently working on it, but it will still take some time.
M3/M4 takes more time...

That’s what you get if the platform holder doesn’t care about bare metal Linux on its computers. Apple makes zero money in configuring its Sillicon to run Linux.

Although, the Asahi team did do they an excellent job with what they had. Also look at Qualcomm’s X Elite, it’s still to this day not fully supported on Linux despite Qualcomm promising support.

x86 is easier to work with when it comes to Linux cause AMD and Intel maintain support.

 

[poke01]

Beyond the Roar: Dissecting Intel’s Panther Lake

Intel is gearing up to launch its Core Ultra 300/400 processors, namely Panther Lake, later this year. In this article, I’ll uncover…

medium.com

This is a nice article to read, like he mentions it will be interesting to see Panther lake cover <15 watts and >25 watts

 

[Io Magnesso]

That’s what you get if the platform holder doesn’t care about bare metal Linux on its computers. Apple makes zero money in configuring its Sillicon to run Linux.

Although, the Asahi team did do they an excellent job with what they had. Also look at Qualcomm’s X Elite, it’s still to this day not fully supported on Linux despite Qualcomm promising support.

x86 is easier to work with when it comes to Linux cause AMD and Intel maintain support.

In a way, Intel/AMD, which contributes well to open source, is amazing...
I understand that you can still run a consumer product that can be used for embedded and edges to be able to run on Linux.
Even complete consumer products like Lunar Lake that aren't very suitable for incorporation or edges I'm glad that it works fine on Linux.

 

[DrMrLordX]

Ampere Altra Max Continues To Deliver Competitive Power Efficiency To AMD EPYC & Intel Xeon - Phoronix

www.phoronix.com

It kinda looks like (which is why I specifically used that wording) ARM does not do well in the current state of servers.

Ampere's products from 2023 have nothing to do with M4. Which in turn has little to do with this thread . . .

 

[511]

Well when Apple gives backwards compatibility over 20 years with the additional features of RAS Confidential computing and other stuff without breaking stuff they can enter DC.

 

[poke01]

Well when Apple gives backwards compatibility over 20 years with the additional features of RAS Confidential computing and other stuff without breaking stuff they can enter DC.

Nvidia is not doing that and they entered DC. Anyway, I don't even know how DC became a topic in a client focused thread.

 

[511]

Nvidia is not doing that and they entered DC. Anyway, I don't even know how DC became a topic in a client focused thread.

They are supporting older hardware and don't jump ISAa like Apple they provide a good platform for developers.
For 2nd part it always happens in Intel/AMD thread things going off topic. 🤣🤣

 

[OneEng2]

I guess I didn't realize that all of the subject matter in the thread is client.

I still think that the ability of Intel to make client chips will be dependent on their ability to be successful in the DC market.

 

[511]

They need a good core architecture first for both client and server most of their problem stems from that.

 

[dullard]

I guess I didn't realize that all of the subject matter in the thread is client.

I still think that the ability of Intel to make client chips will be dependent on their ability to be successful in the DC market.

We do have a thread specifically for Intel Rapids products, right on the front page of the CPU forum. https://forums.anandtech.com/threads/intel-current-and-future-lakes-rapids-thread.2509080/ This thread started off as Meteor Lake which is client only. It slowly expanded to include other client CPUs.

It will soon be time for this thread to die though. After Panther Lake launches we should let this one go and move on to the Nova Lake thread.