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.



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	8 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 (20A)	Arrow Lake (N3B)	Arrow Lake Refresh (N3B)	Lunar Lake	Panther Lake
Platform	Mobile H/U Only	Desktop Only	Desktop & Mobile H&HX	Desktop Only	Mobile U Only	Mobile H
Process Node	Intel 4	Intel 20A	TSMC N3B	TSMC N3B	TSMC N3B	Intel 18A
Date	Q4 2023	Q1 2025 ?	Desktop-Q4-2024 H&HX-Q1-2025	Q4 2025 ?	Q4 2024	Q1 2026 ?
Full Die	6P + 8P	6P + 8E ?	8P + 16E	8P + 32E	4P + 4E	4P + 8E
LLC	24 MB	24 MB ?	36 MB ?	?	8 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.)

 

[SiliconFly]

It's IOSF. From 2012.
You can even like look for MTL hotchips slideware.

Intel slides during launch said it's NOC. Even anandtech says NOC. Many other say it's NOC. If you believe otherwise, it's perfectly okay. It's a free world and we all have our right to believe what we want. Just for you though...

Link1
Link2
Link3
Link4
Link5

Sometimes we just have to accept the fact that ARL is gonna be superior to Zen5 in some aspects.

I'll repeat it once again esp. for you though: As of now, only LNC appears to be the weak link in ARL. As far as packaging is concerned, ARL is far superior than Zen5.

 

[adroc_thurston]

Intel slides during launch said it's NOC

Yeah IOSF is a NOC.
From 2012. It's for phones.

 

[igor_kavinski]

Yeah IOSF is a NOC.
From 2012. It's for phones.

If that's true, God forgive me if I die from uncontrollable laughter!

 

[SiliconFly]

Yeah IOSF is a NOC.
From 2012. It's for phones.

Yeah. Feel free to believe in IOSF. And that rotary phone & mobile phone tech too. Interesting.

 

[igor_kavinski]

I hope we see some OEM release a mini-ITX or micro-ATX mobo with soldered MTL that can be overclocked with liquid cooling.

Would be interesting to see the actual limits of Intel 4.

 

[ondma]

Intel slides during launch said it's NOC. Even anandtech says NOC. Many other say it's NOC. If you believe otherwise, it's perfectly okay. It's a free world and we all have our right to believe what we want. Just for you though...

Link1
Link2
Link3
Link4
Link5

Sometimes we just have to accept the fact that ARL is gonna be superior to Zen5 in some aspects.

I'll repeat it once again esp. for you though: As of now, only LNC appears to be the weak link in ARL. As far as packaging is concerned, ARL is far superior than Zen5.

Give it a rest with the packaging hype. Dont get me wrong, I would love to see ARL be very competitive or even beat Zen5. It is the final performance and power consumption that counts though. That is the sum total of all parts of the chip. So what if Foveros is superior to AMD's solution? If the core is not as good, that is like putting higher speed rated tires on a car with a less powerful engine. Yes, it may be better, but it is not the determining factor in final performance.

 

[SiliconFly]

Give it a rest with the packaging hype. Dont get me wrong, I would love to see ARL be very competitive or even beat Zen5. It is the final performance and power consumption that counts though. That is the sum total of all parts of the chip. So what if Foveros is superior to AMD's solution? If the core is not as good, that is like putting higher speed rated tires on a car with a less powerful engine. Yes, it may be better, but it is not the determining factor in final performance.

Dont get me wrong either. I just want to put an end to the hype that Zen5 is already better than ARL in all aspects and ARL doesn't even stand a chance at all. It's not true.

If LNC performs admirably, ARL might be very competitive to Zen5 (considering all other components in ARL are pretty much excellent).

 

[Khato]

For those interested, the Meteor Lake Hot Chips presentation - https://hc34.hotchips.org/assets/pr...ake_Hotchips _ Wilfred _ final_submit (1).pdf - slide 26 specifically states that the IOE is using IOSF. This correlates to the "IO Fabric" in the press slide deck - https://images.anandtech.com/doci/20046/SoC and IO Deep Dive_FINAL CLEAN-20.png - not the NOC. Why exactly is there an assumption that Intel's limited to a single protocol?

 

[adroc_thurston]

Why exactly is there an assumption that Intel's limited to a single protocol?

every single fabric is a whole zoo of different protocols, we're talking physical implementation which is yes, IOSF.
Like in 2012 Intel-based phones.
Or Avoton.

Zen5 is already better than ARL in all aspects

Yea.

ARL doesn't even stand a chance at all.

yea.

If LNC performs admirably,

It's more stuff.
More stuff isn't what Intel needs in their big cores.

 

[Khato]

every single fabric is a whole zoo of different protocols, we're talking physical implementation which is yes, IOSF.
Like in 2012 Intel-based phones.
Or Avoton.

IOSF is the protocol used for the connection to the IOE, and internally in the SOC tile for a number of the IP blocks.

The CPU tile and GPU tile are using IDI and iCXL protocols, respectively. Pretty clearly spelled out in the hot chips meteor lake presentation.

 

[SiliconFly]

For those interested, the Meteor Lake Hot Chips presentation - https://hc34.hotchips.org/assets/program/conference/day2/Mobile and Edge/Meteor_Lake_Hotchips _ Wilfred _ final_submit (1).pdf - slide 26 specifically states that the IOE is using IOSF. This correlates to the "IO Fabric" in the press slide deck - https://images.anandtech.com/doci/20046/SoC and IO Deep Dive_FINAL CLEAN-20.png - not the NOC. Why exactly is there an assumption that Intel's limited to a single protocol?

Agree.

tCPU <---> SoC (NoC), tGPU <---> SoC (NoC), SoC <---> IoE (IOSF).

NOC Interconnect Fabric is shown in dark blue.
IO Fabric (IOSF) is shown in red.

Apparently, its multi protocol.

 

[adroc_thurston]

IOSF is the protocol used for the connection to the IOE, and internally in the SOC tile for a number of the IP blocks.

It's all IOSF, that's their point to point distributed NOC.
The I/O part is just non-coherent slaves.
Protocols are whatever, iCXL replaced GTI and that wasn't the big change.

 

[SiliconFly]

It's all IOSF, that's their point to point distributed NOC.
The I/O part is just non-coherent slaves.
Protocols are whatever, iCXL replaced GTI and that wasn't the big change.

NoC replaces point-to-point links with a router-based packet switching network between the CPU, GPU & SoC tiles. Acts more like a physical layer which supports other protocols running on top of its stack.

 

[adroc_thurston]

NoC replaces point-to-point links with a router-based packet switching network between the CPU, GPU & SoC tiles. Acts more like a physical layer which supports other protocols running on top of its stack.

Thank you, chatGPT.

 

[SiliconFly]

Thank you, chatGPT.

Well, I ain't gonna resort to name calling! Sounds too miserable.

 

[Khato]

It's all IOSF, that's their point to point distributed NOC.
The I/O part is just non-coherent slaves.
Protocols are whatever, iCXL replaced GTI and that wasn't the big change.

No, it's not all IOSF. If you actually had access to IOSF specifications you'd know this.

 

[DrMrLordX]

It's not true.

Based on what? Interconnect?

 

[igor_kavinski]

No matter how much time anyone wants to waste debating how excellent the packaging or any other tech Intel has at their disposal, they are determined to not compete as long as they can because they simply do not want to spend more and keep costs down thanks to the mess created by Pat's predecessors.

They could:

increase cache sizes

decrease cache latencies, even by a single cycle at every level

improve their memory controller for lowest latency timings and highest memory speeds

widen their internal I/O datapaths/buses to transfer more data at higher speed

use huge eDRAM or HBM L4 cache

have a future kickass architecture backported to their most refined process node as a plan B

All of the above are very expensive things to do for them in their current financial condition.

If Arrow Lake still manages to impress, it will be a miracle that Intel badly needs.

 

[Hitman928]

First MSI Claw (their MTL based handheld) previews are out, https://www.tomshardware.com/video-...nst-amds-phoenix-apu-in-early-msi-claw-review

They call them reviews, but since it is still not a retail release, I prefer to call them previews. Performance is not good against the Rog Ally handheld built around AMD's Z1 Extreme processor. It is quite a bit slower in every game they tested. Not sure if its down to being strictly power limited compared to the laptops we've seen so far or beta firmware/drivers that need improving (probably both). Hopefully firmware/driver updates can help improve its performance and at least close the gap a bit.

 

[SiliconFly]

Based on what? Interconnect?

It's there in the next statement.

 

[DrMrLordX]

Meteor Lake seems to struggle in gaming benchmarks.

@SiliconFly

That's an "if". A big "if", since the only preview we have for Arrow Lake is that leaked Intel slide showing it beating Raptor Lake by a lousy 20%, at a 251W PL2. Based on that, I'd say Arrow Lake is going to be neck-and-neck with Raptor Lake, just at a lower overall power consumption (unless Intel is dumb enough to ship more CPUs with high PL values).

 

[CouncilorIrissa]

First MSI Claw (their MTL based handheld) previews are out, https://www.tomshardware.com/video-...nst-amds-phoenix-apu-in-early-msi-claw-review

They call them reviews, but since it is still not a retail release, I prefer to call them previews. Performance is not good against the Rog Ally handheld built around AMD's Z1 Extreme processor. It is quite a bit slower in every game they tested. Not sure if its down to being strictly power limited compared to the laptops we've seen so far or beta firmware/drivers that need improving (probably both). Hopefully firmware/driver updates can help improve its performance and at least close the gap a bit.

While I believe that MTL will end up a bit slower, I'd like to point out that ROG Ally is a mature device at this point. It's performance at low power envelopes has gotten a lot better since release, so I think there's quite a bit of headroom for the Claw to grow.

 

[CouncilorIrissa]

LNL's here (presumably):

抱歉，该内容已被作者删除 - 知乎

知乎，中文互联网高质量的问答社区和创作者聚集的原创内容平台，于 2011 年 1 月正式上线，以「让人们更好的分享知识、经验和见解，找到自己的解答」为品牌使命。知乎凭借认真、专业、友善的社区氛围、独特的产品机制以及结构化和易获得的优质内容，聚集了中文互联网科技、商业、影视、时尚、文化等领域最具创造力的人群，已成为综合性、全品类、在诸多领域具有关键影响力的知识分享社区和创作者聚集的原创内容平台，建立起了以社区驱动的内容变现商业模式。

www.zhihu.com

 

[jpiniero]

What a dumb idea to use Meteor Lake in a handheld. It's most likely graphics drivers and CPU power consumption as to the culprit.

 

[FlameTail]

What a dumb idea to use Meteor Lake in a handheld. It's most likely graphics drivers and CPU power consumption as to the culprit.

THEY SHOULDVE WAITED FOR KUBAR LAKLBE