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

[Tigerick]

Wildcat Lake (WCL) Preliminary Specs

Intel Wildcat Lake (WCL) is upcoming mobile SoC replacing ADL-N. 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 Q2/Computex 2026. In case people don't remember AlderLake-N, I have created a table below to compare the detail specs of ADL-N and WCL. Just for fun, I am throwing LNL and upcoming Mediatek D9500 SoC.

		Intel Alder Lake - N	Intel Wildcat Lake	Intel Lunar Lake		Mediatek D9500
Launch Date		Q1-2023	Q2-2026 ?	Q3-2024		Q3-2025
Model		Intel N300	?	Core Ultra 7 268V		Dimensity 9500 5G
Dies		2	2	2		1
Node		Intel 7 + ?	Intel 18-A + TSMC N6	TSMC N3B + N6		TSMC N3P
CPU		8 E-cores	2 P-core + 4 LP E-cores	4 P-core + 4 LP E-cores		C1 1+3+4
Threads		8	6	8		8
Max Clock		3.8 GHz	?	5 GHz
L3 Cache		6 MB	?	12 MB
TDP		7 W	Fanless ?	17 W		Fanless
Memory		64-bit LPDDR5-4800	64-bit LPDDR5-6800 ?	128-bit LPDDR5X-8533		64-bit LPDDR5X-10667
Size		16 GB	?	32 GB		24 GB ?
Bandwidth			~ 55 GB/s	136 GB/s		85.6 GB/s
GPU		UHD Graphics		Arc 140V		G1 Ultra
EU / Xe		32 EU	2 Xe	8 Xe		12
Max Clock		1.25 GHz		2 GHz
NPU		NA	18 TOPS	48 TOPS		100 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.

 

[DavidC1]

From geeker wanView attachment 108191

Ahh, the joys of mobile benchmarking.

@SiliconFly Look, your hated Redwood Cove is 6% faster than Zen 4!

This is why the perf/clock debate will NOT be solved until Arrowlake desktop is here.

@Abwx Those are desktop chips and Geekerwan is using laptop chips doofus.

 

[511]

Lunarlake: https://www.laptopmag.com/laptops/w...s-with-an-astounding-18-hours-of-battery-life

Versus competing chips:

Laptop	Battery life (web surfing hh:mm)	UL Procyon Office Productivity Battery LIfe

Dell XPS 13 9345 (Snapdragon X Elite)	19:01	Not tested
Dell XPS 13 9350 (Intel Core Ultra 7 258V)	18:34	23:38
Apple MacBook Pro 14 M3	17:16	Not tested

Why I wouldn't get an OLED display:

Laptop	Battery life (web surfing hh:mm)	UL Procyon Office Productivity Battery LIfe
Dell XPS 13 9350 (Intel Core Ultra 7 258V)	18:34	23:38
Dell XPS 13 9350 OLED (Intel Core Ultra 7 258V)	08:40	12:48

Versus other Lunarlake device:

Laptop	Battery life (web surfing hh:mm)	UL Procyon Office Productivity Battery LIfe
Asus Zenbook S 14 (Intel Core Ultra 7 258V)	13:51	16:21
Dell XPS 13 9350 OLED (Intel Core Ultra 7 258V)	08:40	12:48

Same reason I don't buy oled until it's better you turn a white image or crank brightness and it will destroy battery like nothing else something even apple is afraid of on laptops

 

[DavidC1]

Same reason I don't buy oled until it's better you turn a white image or crank brightness and it will destroy battery like nothing else something even apple is afraid of on laptops

Worse, OLED will have burn in issues. Why would I pay $1800 dollars for a "premium" laptop with a display with worse lifespan than a QLC SSD without DRAM cache and cut battery life anywhere from 33 to 50%?

OLED displays are -$300 worth to me.

 

[9949asd]

From geeker wanView attachment 108191

It’s not. This Notebook ba lab.

 

[igor_kavinski]

From geeker wanView attachment 108191

Very cool. So Zen 5c and Skymont are evenly matched.

 

[Abwx]

@Abwx Those are desktop chips and Geekerwan is using laptop chips doofus.

This change nothing since it s Spec_int, beside Zen 4 mobile is the same as Zen 4 DT, and it cant be just a few % better than a Zen 2, be it a DT one, there s at least 25-30% INT IPC between those gens.

 

[DavidC1]

This change nothing since it s Spec_int, beside Zen 4 mobile is the same as Zen 4 DT, and it cant be just a few % better than a Zen 2, be it a DT one, there s at least 25-30% INT IPC between those gens.

You might want to take notice of the fact that Intel chips there aren't performing well in that test either. But I know you don't want to.

Very cool. So Zen 5c and Skymont are evenly matched.

Yup, radically different approach, company, process and the results are similar.

This is the key though. An actual efficiency core.

 

[Abwx]

You might want to take notice of the fact that Intel chips there aren't performing well in that test either. But I know you don't want to.

Yup, radically different approach, company, process and the results are similar.

Actualy the 155H perform better at Geekerwan, by 7.4%, than at Chip and Cheese who use a more recent GCC compiler, so what is to be aknowledged here.?.

His Zen 5 score is 8.8 while if i downscale CaC score to 5.1Ghz i end at 10.647, that is, 21% higher than his alleged score at 5.1GHz.

Likewise his Zen 4 score is 7.99 at 5.2 Ghz while a downscale of CaC score yield 9.579, that is, almost 20% difference, and curiously, but unsurprisingly, the same as with Zen 5.

 

[9949asd]

The best of lunar lake is gpu.

 

[511]

Worse, OLED will have burn in issues. Why would I pay $1800 dollars for a "premium" laptop with a display with worse lifespan than a QLC SSD without DRAM cache and cut battery life anywhere from 33 to 50%?

OLED displays are -$300 worth to me.

That as well the manufacturers can market all they want but Oled Lifespan is something different also on Mobile phones they usually go to sleep that prevents burn in but on laptop people treat it as a desktop it stays at the same picture for hours we disable sleep and other things that we don't do with Mobile phones
A gamer example we leave steam/torrent on download for a large file it is mostly static picture excluding small area overnight

 

[DavidC1]

For those of you that think Intel's P core team is still good, David Huang points out in his review that the branch mispredict in Lion Cove regressed compared to Golden Cove.

In contrast, Skymont's branch predictor improved substantially over Gracemont, and now exceeds not only Zen 4's MKPI performance but also Lion Cove's as well.

Huang also points out that the performance per watt of Lion Cove cluster is poor despite being on a new process.

P core team, it's time to be replaced!

That as well the manufacturers can market all they want but Oled Lifespan is something different also on Mobile phones they usually go to sleep that prevents burn in but on laptop people treat it as a desktop it stays at the same picture for hours we disable sleep and other things that we don't do with Mobile phones

I simply don't want to worry about my display flunking. I had desktop monitors have problems after 5 years of usage(Samsung models) but that's regular degradation. OLED burn-in is pretty much a certainty.

OLED is perfect for modern throw-away junk that needs borrowing credit to afford it.

 

[DavidC1]

Intel absolutely nailed Skymont based on that power/perf graph by Huang. The performance over Crestmont is lot lower than expected but makes up by the stunning improvements in power efficiency. I have to think they are probably related, and Arrowlake's Skymont will have a different perf/watt behavior. Skymont is really the star here in driving Lunarlake's efficiency probably all the way to moderate usage.

It beats everything to quite high power levels but it can truly scale down way low.

On a side note, you can see how MTL's LPE core power numbers are seriously hobbled by bad design, because powering down uncore makes power efficiency similar to E cores, while on default mode it is much less efficient.

 

[511]

For those of you that think Intel's P core team is still good, David Huang points out in his review that the branch mispredict in Lion Cove regressed compared to Golden Cove.

In contrast, Skymont's branch predictor improved substantially over Gracemont, and now exceeds not only Zen 4's MKPI performance but also Lion Cove's as well.

Huang also points out that the performance per watt of Lion Cove cluster is poor despite being on a new process.

P core team, it's time to be replaced!

We know for the past few archs that it is bad Skymont is just amazing improvement P core PPA is dead they need to redesign it from ground

I simply don't want to worry about my display flunking. I had desktop monitors have problems after 5 years of usage(Samsung models) but that's regular degradation. OLED burn-in is pretty much a certainty.

OLED is perfect for modern throw-away junk that needs borrowing credit to afford it.

This as well IPS doesn't have the shit reliability of oled at least mini led is the best of both world

 

[DavidC1]

We know for the past few archs that it is bad Skymont is just amazing improvement P core PPA is dead they need to redesign it from ground

It's the TEAM in P core that needs to be dismantled and rebuilt from the ground.

Because if it was just the design, at least they could fix it. This kind of improvement after 3 years just piss-poor.

This as well IPS doesn't have the shit reliability of oled at least mini led is the best of both world

I remember when OLED was said to be "organic" and that it could lower power drastically because it doesn't need a backlight and I was young at that time and hoped for it but had bit of skepticism.

Some things don't come for free. You can't have the best contrast, best color and have lower power consumption. Something has to give. Yea you can do little better with better engineering, that always exist.

 

[jdubs03]

Supposedly, that’s where Nova Lake and its mobile contemporary supposed to come in with the redesigned p-core. Panther/coyote cove to the rescue!

 

[DavidC1]

Supposedly, that’s where Nova Lake and its mobile contemporary supposed to come in with the redesigned p-core. Panther/coyote cove to the rescue!

Yea, I hope the E core team leads that.

 

[igor_kavinski]

It's the TEAM in P core that needs to be dismantled and rebuilt from the ground.

Hard to do. It's IDC, from what I understand and they have Intel by the balls with their claim that Intel owes them for getting them out of the P4 mess with Banias and later Core arch. Remember, Intel is a haven of dirty politics.

 

[DavidC1]

Hard to do. It's IDC, from what I understand and they have Intel by the balls with their claim that Intel owes them for getting them out of the P4 mess with Banias and later Core arch. Remember, Intel is a haven of dirty politics.

The IDC of up to 5 years post-Core is not the IDC today. BK basically gutted all the important people.

It would be very unfortunate if they can't change that, because IDC at the PEAK was nowhere near the pace of development the current E core team is on.

According to reports on reddit they were already getting greedy by Ivy Bridge era. Take it with a grain of salt, but the execs did not see the point investing in uarch because AMD was uncompetitive, and they had no need to go beyond 4 cores when AMD had chips such as Bulldozer.

Remember how disappointing Lion Cove is with the 10-12% gain? Well That's what we got with Haswell and Skylake.

 

[511]

Hard to do. It's IDC, from what I understand and they have Intel by the balls with their claim that Intel owes them for getting them out of the P4 mess with Banias and later Core arch. Remember, Intel is a haven of dirty politics.

Yes the Politics something no one at Intel can control and one of the reason for the mess they are in the politics were under the CEOs control for long until Paul/BK took over 🤣

 

[H433x0n]

The 128C Xeon has a 11.8% advantage at 39% more power than a 96C Zen 4, that put things in perspective, at same 500W power the 96C Epyc would outmatch the 128C Xeon.

With 33% more cores the Zen 5 based Epyc will have 1.33 x 1.16 = 54% higher throughput than the 96C Genoa if TDP is increased to 500W up from the curent 360W,
and enough to outmatch the Xeon by 35% at same power or eventualy have the same perfs but then at 0.5x the power.

And there s the dense version that wil add 18% perf/wat on top of the numbers above, that is to ay that Intel wil stil be late by a full, if not two, gens.



He better use a more recent version of GCC because his numbers for Zen 4 and 5 do not make sense at all, per clock his Zen 5 score is lower than the one of a Zen 4 and his Zen 4 is just slightly better than a Zen 2, but the score of the 155H for sure is not underestimated...

Running SPEC CPU2017 at Chips and Cheese?

SPEC, or Standard Performance Evaluation Corporation, maintains and publishes various benchmark suites that are often taken as an industry standard.

chipsandcheese.com

It’s 25% and 22% stronger than traditional Genoa and Bergamo respectively. The CPU you’re referencing is Genoa vcache SKU, which I don’t think is going to be available at the initial launch on October 10th.

 

[H433x0n]

For those of you that think Intel's P core team is still good, David Huang points out in his review that the branch mispredict in Lion Cove regressed compared to Golden Cove.

In contrast, Skymont's branch predictor improved substantially over Gracemont, and now exceeds not only Zen 4's MKPI performance but also Lion Cove's as well.

Huang also points out that the performance per watt of Lion Cove cluster is poor despite being on a new process.

P core team, it's time to be replaced!

I would’ve liked more from Lion Cove as well but you’re losing the forest for the trees. There’s only 1 other x86 competitor and against Zen 5 it’s just as performant (if not a few percent better in integer performance) while not drawing any extra power.

If somebody said that would be the result 6 months ago nobody on this forum would’ve believed you.

 

[igor_kavinski]

Remember how disappointing Lion Cove is with the 10-12% gain? Well That's what we got with Haswell and Skylake.

Skylake was the beginning of the end for Intel. The bugs in that were what prompted Apple to ditch them coz Apple engineers were basically doing Intel's job for them by finding silicon bugs and reporting them to Intel. And Apple still gave them business till Ice Lake. I like to think that Skylake was when they decided to lie on their back and ass and all they could see was the sky, rather than incoming traffic. Pat tried to steer them but in the wrong direction!

 

[DrMrLordX]

Yes it will require different mask

New masks required.

Okay I thought so. So on to my next question: Intel didn't just buy N3B from TSMC's existing production capacity. Intel paid up front for an entire fab to be built to fill their order. TSMC was also forced to refit a research fab to full production to complete the order. Again, the devil's in the details (e.g. the contract), but wouldn't facilities built for N3B production require retooling before they could run N3E/N3P? Or am I overthinking this? Yes I suppose Intel could take N3E/N3P wafers per their contract from other facilities assuming TSMC is flexible enough to fill those orders, but . . .

Falcon Shores hasn't been cancelled as far as I know, only the CPU+GPU version was cancelled, the GPU only version is still on the roadmap.

This article is the closest I've seen to recent news regarding Falcon Shores:

Intel’s Falcon Shores Future Looks Bleak as It Concedes AI Training to GPU Rivals

Intel’s Falcon Shores future looks bleak as it concedes AI training to GPU rivals On Monday, Intel sent a letter to employees detailing its comeback plan after an abysmal second-quarter […]

www.hpcwire.com

Technically there was no official announcement of Falcon Shores being cancelled, though Gelsinger seemed to imply that it was DoA due to his statement about not competing with NV and AMD in the GPU AI training space.

 

[controlflow]

OLED vs LCD battery life. Ouch...

Dell XPS 13 9350 + Lunar Lake Review: Intel’s Back With A Vengeance

Astounding battery life and massive leaps forward in CPU and GPU performance prove that Intel is done messing around. Click here for Lunar Lake benchmarks.

www.forbes.com

 

[jur]

Technically there was no official announcement of Falcon Shores being cancelled, though Gelsinger seemed to imply that it was DoA due to his statement about not competing with NV and AMD in the GPU AI training space.

Even AMD's GPUs are essentially DoA for training. Nvidia has massive software, interconnect / scale advantage. The new compute unit for them are racks with massive interconencts between the gpu. Google is the only one (as far as I know) that has similar interconnect infrastructure. AMD and Intel can only compete at smaller model traning and inference. We'll see what happens when Intel starts integrating silicon photonics into their hardware, but this will take years. Intel at least has solid software compared to AMD.