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

[Tigerick]

Wildcat Lake (WCL) Specs

Intel Wildcat Lake (WCL) is upcoming mobile SoC replacing Raptor Lake-U. WCL consists of 2 tiles: compute tile and PCD tile. It is true single die consists of CPU, GPU and NPU that is fabbed by 18-A process. Last time I checked, PCD tile is fabbed by TSMC N6 process. They are connected through UCIe, not D2D; a first from Intel. Expecting launching in Q1 2026.

		Intel Raptor Lake U	Intel Wildcat Lake 15W?	Intel Lunar Lake		Intel Panther Lake 4+0+4
Launch Date		Q1-2024	Q2-2026	Q3-2024		Q1-2026
Model		Intel 150U	Intel Core 7	Core Ultra 7 268V		Core Ultra 7 365
Dies		2	2	2		3
Node		Intel 7 + ?	Intel 18-A + TSMC N6	TSMC N3B + N6		Intel 18-A + Intel 3 + TSMC N6
CPU		2 P-core + 8 E-cores	2 P-core + 4 LP E-cores	4 P-core + 4 LP E-cores		4 P-core + 4 LP E-cores
Threads		12	6	8		8
Max Clock		5.4 GHz	?	5 GHz		4.8 GHz
L3 Cache		12 MB		12 MB		12 MB
TDP		15 - 55 W	15 W ?	17 - 37 W		25 - 55 W
Memory		128-bit LPDDR5-5200	64-bit LPDDR5	128-bit LPDDR5x-8533		128-bit LPDDR5x-7467
Size		96 GB		32 GB		128 GB
Bandwidth				136 GB/s
GPU		Intel Graphics	Intel Graphics	Arc 140V		Intel Graphics
RT		No	No	YES		YES
EU / Xe		96 EU	2 Xe	8 Xe		4 Xe
Max Clock		1.3 GHz	?	2 GHz		2.5 GHz
NPU		GNA 3.0	18 TOPS	48 TOPS		49 TOPS

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.

 

[Hitman928]

Are you denying the part in bold, that the 18A process is on schedule? Or did you interpret that incorrectly as Panther Lake is a mid-2025 release?

TPU kind of butchered it, but If you read the direct quote from Pat Gelsinger, I don't think I'm misinterpreting what he was promising. You could also just read the title of the linked TPU article to see what he was saying, but here's the direct quote for you:

"We expect to release the 1.0 PDK for Intel 18A this quarter. Furthermore, our lead products, Clearwater Forest and Panther Lake are already in fab, and we expect to begin production ramp of Intel 18A in these products in the first half of '25 for product release in the middle of next year."

Pat Gelsinger - Intel CEO

 

[dullard]

TPU kind of butchered it, but If you read the direct quote from Pat Gelsinger, I don't think I'm misinterpreting what he was promising. You could also just read the title of the linked TPU article to see what he was saying, but here's the direct quote for you:

"We expect to release the 1.0 PDK for Intel 18A this quarter. Furthermore, our lead products, Clearwater Forest and Panther Lake are already in fab, and we expect to begin production ramp of Intel 18A in these products in the first half of '25 for product release in the middle of next year."

Pat Gelsinger - Intel CEO

That link better supports your idea.

But, (1) it is well more than a year older than most of the links that I shared. It is quite forward-thinking. In fact, the article also covers 20A which was cancelled. Things change in 1.5 years for technology. Heck the CHIPS act funds were delayed by about that much time. Yes, 18A yields aren't where they want them to be so they did not launch products in mid 2025.

And (2) still doesn't actually say that Panther lake will be launched in volume in 2025. It just says "product release".

 

[regen1]

Intel makes the CPUs, not the laptops. When Intel launches, sells, and ships a laptop CPU that has almost nothing to do with when a laptop manufacturer launches their laptops. As DavidC1 said above, desktops generally just plop the CPU on in and you ship it. Laptops take months to optimize after you have the CPUs.

Well, Intel doesn't make laptop that's correct but Intel does ship it out and has to be in schedule for a launch(now one can claim Intel doesn't control OEM's launch, etc.).
All you have to do is look at Meteor Lake's launch for reference(the EEP with that, Zenbook 14, extremely low availability in 2023). If there was no such hurry they could have avoided the December 2023 event and the EEP that year, could have waited less than a month for CES 2024.

In the end it really doesn't matter for high-volume availability whether you have something the December of previous year which is extremely limited in availability and not accessible in all regions that easily.

 

[regen1]

Any info about Wild Cat Lake?
I am still thinking that giving HT on those processors could help them.

Same as what was already known
2P + 0E + 4LPE , 2Xe3 iGPU, NPU: 18TOPS(?)
Integrated Thunderbolt 4

WCL doesn't have HT and why would the cores in Wildcat Lake have HT when the same cores used in Panther Lake don't ?

 

[dullard]

All you have to do is look at Meteor Lake's launch for reference(the EEP with that, Zenbook 14, extremely low availability in 2023). If there was no such hurry they could have avoided the December 2023 event and the EEP that year, could have waited less than a month for CES 2024.

That is just how Intel operates. Meteor Lake isn't the only example.

Look at Alder Lake too. It was launched Oct 27, 2021 but the cheaper desktops weren't available until January 2022 and the mobile launch wasn't complete until May 2022.

Raptor Lake: launched Oct 20, 2022. But most of the chips were launched at CES in January 2023.

 

[Hitman928]

That link better supports your idea.

But, (1) it is well more than a year older than most of the links that I shared. It is quite forward-thinking. In fact, the article also covers 20A which was cancelled. Things change in 1.5 years for technology. Heck the CHIPS act funds were delayed by about that much time. Yes, 18A yields aren't where they want them to be so they did not launch products in mid 2025.

And (2) still doesn't actually say that Panther lake will be launched in volume in 2025. It just says "product release".

Your attempted framing of this is entertaining. The actual point is that Intel has a long history now of knowingly over promising and under delivering / missing targets so it's very natural for people to be skeptical. I will say that LBT seems to be much more honest/transparent than Pat was but Intel needs to re-establish its reputation before people stop looking for discrepancies and pointing out potential ambiguities in their timelines/promises.

 

[dullard]

The actual point is that Intel has a long history now of knowingly over promising and under delivering / missing targets so it's very natural for people to be skeptical.

That point is true. Intel does overpromise. But, Intel still has been consistent about Panther Lake's launch. Their first SKU (or sometimes they said "relatively few products") in 2025 and the bulk in 2026. That too could be an overpromise--but it has at least been consistent. There has never been a time to actually think you'd get much Panther Lake in 2025 unless you were incorrectly interpreting what Intel said.

 

[regen1]

That is just how Intel operates. Meteor Lake isn't the only example.

Look at Alder Lake too. It was launched Oct 27, 2021 but the cheaper desktops weren't available until January 2022 and the mobile launch wasn't complete until May 2022.

Raptor Lake: launched Oct 20, 2022. But most of the chips were launched at CES in January 2023.

That's a little different. Regular Desktop SKUs launch vs Laptop only series launch.
Desktop chips launches have been like this for sometime : Sept/Oct announcement(K-series SKUs) and availability in Q4(October).
Non-K series announced in CES(Jan) and then subsequent availability.
MTL launch was different. It definitely was a somewhat hurried launch(It was a showcase and first thing for Intel 4 and some projections(timeline) regarding it).
PTL is similar for 18A.
For ArrowLake laptop SKUs there was no hurry(no such projections were made, not needed), got launched in CES 2025.

 

[Hitman928]

That point is true. Intel does overpromise. But, Intel still has been consistent about Panther Lake's launch. Their first SKU (or sometimes they said "relatively few products") in 2025 and the bulk in 2026. That too could be an overpromise--but it has at least been consistent. There has never been a time to actually think you'd get much Panther Lake in 2025 unless you were incorrectly interpreting what Intel said.

So when Pat said it would launch mid 2025, people should not have expected to be able to have at least a few models to choose from by Christmas? So are you not expecting multiple models available until autumn 2026 now?

 

[511]

For ArrowLake laptop SKUs there was no hurry(no such projections were made, not needed), got launched in CES 2025.

ARL Laptop launched Q1 25 same timeframe as PTL volume launch

 

[regen1]

ARL Laptop launched Q1 25 same timeframe as PTL volume launch

That's known but don't ignore the aspects of (i)no process node timelines attached to ARL laptop SKUs vs MTL(Intel 4) and PTL's(Intel 18A)
(ii)no projections of ARL laptop SKUs was made to be launched in 2024.

I think you know the schedule of MTL launch. Would MTL EEP/launch be required that much early when CES was just a few weeks away ?
For reference Meteor Lake got launched in Dec 14, 2023 with very little availability in 2023(esp. considering worldwide) via EEP.
CES 2024 was Jan9-12, 2024.

Anyway I don't think it's that big an issue. But there has been a subtle difference in launch that could be acknowleded even though for most part(high-volume) it doesn't matter.

 

[511]

As long as Intel recognizes PTL Revenue and They ship PTL chips to OEM they met their commitments.

 

[dullard]

So when Pat said it 18A would launch mid 2025, people should not have expected to be able to have at least a few models to choose from by Christmas? So are you not expecting multiple models available until autumn 2026 now?

Fixed that again for you. Pat never said Panther Lake would be mid-2025. He said 18A would launch mid-2025. Reread your links.

I am expecting a Panther Lake announcement late this year with maybe one model. I am not expecting multiple models of Panther Lake until CES (Jan 2026). And it will ramp slowly from there.

18A yields are not high enough to be profitable right now. Even if 18A was profitable, Intel doesn't have the wafers. As in a physical wafer shortage. AI is sucking all wafers up. Only the highest profit CPUs will be made early 2026. Think mostly server chips.

 

[Hitman928]

Fixed that again for you. Pat never said Panther Lake would be mid-2025. He said 18A would launch mid-2025. Reread your links.

I am expecting a Panther Lake announcement late this year with maybe one model. I am not expecting multiple models of Panther Lake until CES (Jan 2026). And it will ramp slowly from there.

18A yields are not high enough to be profitable right now. Even if 18A was profitable, Intel doesn't have the wafers. As in a physical wafer shortage. AI is sucking all wafers up. Only the highest profit CPUs will be made early 2026. Think mostly server chips.

Not sure how I'm misinterpreting this, "begin production ramp of Intel 18A in these products in the first half of '25 for product release in the middle of next year." How should I read that as anything other than exactly what it says?

 

[DrMrLordX]

Hey just a PSA from a regular user that couldn't use AT earlier today/late last night due to an apparent database outage: please refer to my sig (if you can read it), but also there is a thread on forum stabilitiy issues:

Page 16 - Question - so long and thanks for all the fish

Page 16 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

Sorry for the off-topic but there's a lot of good information and users here, and it would be nice to minimize interruptions in the event that the database doesn't come back online next time in a timely fashion, or at all.

Not sure how I'm misinterpreting this, "begin production ramp of Intel 18A in these products in the first half of '25 for product release in the middle of next year." How should I read that as anything other than exactly what it says?

An excellent question, which should be answered in a few months. Panther Lake will get there when it gets there. For Intel's sake, let's hope it doesn't take them until June/July to actually ship in volume, but it may take them awhile to get the entire product stack out there.

 

[Joe NYC]

Not sure how I'm misinterpreting this, "begin production ramp of Intel 18A in these products in the first half of '25 for product release in the middle of next year." How should I read that as anything other than exactly what it says?

This could have also been related to Clearwater Forest, which has now been twice delayed.

Originally, it was to be released in H2 2026 and delayed from H2 2025 to 2026 due to packaging (as we were told) not because 18A.

 

[ToTTenTranz]

When did they bring back the ability of the GPU to share CPU L3 cache? They stopped it with Meteorlake. I don't remember Lunarlake being able to share CPU L2.

iGPU can't use L3 afaik only L2/SLC also PTL Supports LPDDR5X upto 153GB/s

Lunar Lake's 140V can't access the CPU L3, but Panther Lake's iGPU can.

Intel Panther Lake Technical Deep Dive

Panther Lake is Intel’s new mobile architecture built on the 18A process, featuring RibbonFET transistors, PowerVia backside power delivery, and Foveros-S packaging. With new Cougar Cove and Darkmont cores, Xe3 graphics, and smarter AI, it’s Intel’s most advanced design yet—our writeup breaks...

www.techpowerup.com

The unified memory fabric provides a single address space for all major blocks: the CPU cores, GPU tile, IPU, and NPU. Each can access shared data directly, without explicit copy operations or separate synchronization buffers. I asked Intel about the GPU tile's ability to access the cache, and Intel confirmed that the GPU tile can snoop into the L3 cache on the compute tile, allowing it to read or share data directly with the CPU. This cache-level visibility helps avoid redundant memory transfers and improves latency in mixed workloads such as AI-assisted workloads or video processing.

Also, the NPU being able to use the L3 means it shares a cache with the iGPU.
It might be possible to offload FSR4 INT8 or XeSS DP4a to the NPU.

 

[adroc_thurston]

Lunar Lake's 140V can't access the CPU L3, but Panther Lake's iGPU can.

Yeah it can, they all can snoop into CPU caches.
But what differentiated Intel iGPs since Gen6 until Gen12.7 in MTL was the ability to use CPU L3 as gfx LLC. Not the case anymore.

 

[Tigerick]

Same as what was already known
2P + 0E + 4LPE , 2Xe3 iGPU, NPU: 18TOPS(?)
Integrated Thunderbolt 4

WCL doesn't have HT and why would the cores in Wildcat Lake have HT when the same cores used in Panther Lake don't ?

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) - Product Specifications | Intel

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) quick reference with specifications, features, and technologies.

www.intel.com

Does that make you re-think why Intel NOT enabling HT in the rest of ARL lineup ? Oh yeah, this ARL is fabbed by Intel 3, not N3B.

I did not say WCL will have HT enable, but I won't say never...

 

[regen1]

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) - Product Specifications | Intel

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) quick reference with specifications, features, and technologies.

www.intel.com

Does that make you re-think why Intel NOT enabling HT in the rest of ARL lineup ?

I did not say WCL will have HT enable, but I won't say never...

So you don't realise that ARL-U parts are basically MTL-U refreshed(with compute tile on Intel 3 with some clock gains).

MTL, P-cores= Redwood Cove (have HT), E and LPE = Crestmont
ARL-U, P cores= Redwood Cove(have HT), E and LPE = Crestmont

ARL-H ARL-HX ARL-S, P cores = Lion Cove, and no HT in those series

 

[MoistOintment]

Fixed that again for you. Pat never said Panther Lake would be mid-2025. He said 18A would launch mid-2025. Reread your links.

I am expecting a Panther Lake announcement late this year with maybe one model. I am not expecting multiple models of Panther Lake until CES (Jan 2026). And it will ramp slowly from there.

18A yields are not high enough to be profitable right now. Even if 18A was profitable, Intel doesn't have the wafers. As in a physical wafer shortage. AI is sucking all wafers up. Only the highest profit CPUs will be made early 2026. Think mostly server chips.

When 18A launches, when PTL launches (from a revenue POV), and when PTL based laptops become available to purchase are all 3 separate events, months apart from each other.

And I feel like Intel does lean into this to obscure launch windows a bit, and these should be further clarified by them when discussing. As a foundry (even if Intel Design is their only client), the "launch" of 18A is, by definition, months earlier than PTL laptops being on store shelves. But it's very easy to confuse all of these events and I feel like that's Intel's goal with a lot of their announcements.

 

[MoistOintment]

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) - Product Specifications | Intel

Intel® Core™ Ultra 7 Processor 255U (12M Cache, up to 5.20 GHz) quick reference with specifications, features, and technologies.

www.intel.com

Does that make you re-think why Intel NOT enabling HT in the rest of ARL lineup ? Oh yeah, this ARL is fabbed by Intel 3, not N3B.

I did not say WCL will have HT enable, but I won't say never...

SMT isn't just a matter of "enabling". If they were able to launch WCL with SMT, then they'd be able to launch the entire PTL product line with it enabled too.

 

[DavidC1]

It's going to be May-June 2026 for any real launch anyway. Even if you could order it by December, at that point it still takes several weeks at least to arrive, and they aren't all desirable parts. Mobile launches are basically a 6-month long ad campaign before you can get your hands on one. I still flag Tigerlake for 3 paper launches, with barely any new info for each other. Even Hot Chips was "corrupted" and is a marketing tactic for them.

 

[ToTTenTranz]

Yeah it can, they all can snoop into CPU caches.
But what differentiated Intel iGPs since Gen6 until Gen12.7 in MTL was the ability to use CPU L3 as gfx LLC. Not the case anymore.

The "read or share data with the CPU" statement strongly suggests it's a snoop with cache update functionality, i.e. it's write capable.

I'm not sure that L-shaped glue was worth the die area if the coprocessors couldn't write in the L3.

 

[DavidC1]

The "read or share data with the CPU" statement strongly suggests it's a snoop with cache update functionality, i.e. it's write capable.

I'm not sure that L-shaped glue was worth the die area if the coprocessors couldn't write in the L3.

Those points also apply to Lunarlake. Pantherlake uses the same Scalable Fabric Gen 2 as Lunarlake.