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.

 

[poke01]

There may be a possibility that an 8P/16E die will be manufactured with an Intel 18A.
It may be possible to make a multi-die even with the 18A version of the die.
8p/16e Die is It may be produced by both IFS and TSMC.

Pat is that really you?

 

[Io Magnesso]

Pat is that really you?

I'm not a Pat... My icon is just a PAT icon with the face of a member of the Japanese artist "Nona Reeves".
Search for "Enjoyee!" by Nona Reeves
That's the original meme
If I had to say it, in August of last year, I was mistakenly followed by a PAT (official) account on X (Twitter).

 

[Io Magnesso]

it would be nice if Nova Lake and (preferably Zen6) had 4 channels of memory channels.
I want the number of channels to be increased even though there are 4 memory slots that have not changed since before.

 

[511]

it would be nice if Nova Lake and (preferably Zen6) had 4 channels of memory channels.
I want the number of channels to be increased even though there are 4 memory slots that have not changed since before.

Not happening sadly at least with DDR 6 we will have our generational upgrade.

 

[coercitiv]

Adored has been MIA

I'm saying I think this leak will not age well. It implies Intel sticking to their current price floor while spending double on compute tiles from TSMC.

I'll happily take them off their hands for such a bargain price, but somehow I doubt the company that speaks of 50% gross margin red lines is about to hold a firesale with premium TSMC silicon.

 

[511]

I'm saying I think this leak will not age well. It implies Intel sticking to their current price floor while spending double on compute tiles from TSMC.

I'll happily take them off their hands for such a bargain price, but somehow I doubt the company that speaks of 50% gross margin red lines is about to hold a firesale with premium TSMC silicon.

who said they are going to price it at the same as ARL same with Zen6 they will charge a premium for sure.

 

[inquiss]

it would be nice if Nova Lake and (preferably Zen6) had 4 channels of memory channels.
I want the number of channels to be increased even though there are 4 memory slots that have not changed since before.

"I would prefer to make all consumer hardware more expensive" not going to happen

 

[DAPUNISHER]

Any more copy pasta AI posts and we start playing ban bingo along with post deletion.

Mod DAPUNISHER

 

[511]

What did I miss 😅

 

[Josh128]

mmm Im skeptical of it matching 32C of Zen 5 TR in MT. TR has 350W+ of TDP to play with, basically double the power of a 9950X, which itself is 200W (170W TDP) of electrical power at stock. I mean, maybe, if its really 18A and they can pull some impressive perf/w efficiency out of that, but is Intel really that hellbent on dying on the 300W+ on desktop hill?

Even more skeptical it can match 32C Zen 5 TR now that they claim 150W, unless, that 150W is a completely fake and useless number.

Doesnt Intel have an Arrow Lake -S desktop "refresh" coming this year?

 

[511]

Even more skeptical it can match 32C Zen 5 TR now that they claim 150W, unless, that 150W is a completely fake and useless number.

Doesnt Intel have an Arrow Lake -S desktop "refresh" coming this year?

at 150W maybe but at 300W it can tbh

 

[coercitiv]

who said they are going to price it at the same as ARL

The market says so. They can price their premium SKUs as high as they want, they can even invent another segment like Ultra 9 XL, but they're not in a position to radically change pricing for Ultra 7 and Ultra 5. That would take massaging over several generations.

same with Zen6 they will charge a premium for sure.

How are AMD going to charge a premium when you just said Ultra 5 will kill Ryzen 7?

Even more skeptical it can match 32C Zen 5 TR now that they claim 150W, unless, that 150W is a completely fake and useless number.

The 285K is 125W TDP, with 250W "Maximum Turbo Power". A base power of 150W would indicate a Turbo Power of 300W+. Remember the 14900KS has an official Intel Extreme profile of 320W, sustained. Officially, the 14900KS has a TDP of 150W

 

[AcrosTinus]

Even more skeptical it can match 32C Zen 5 TR now that they claim 150W, unless, that 150W is a completely fake and useless number.

Doesnt Intel have an Arrow Lake -S desktop "refresh" coming this year?

If you know Intel, this 150W are the locked down, not tuned, basically constrained specs. Basically a PR move to ward off the techtuber crowd that will bench a 8C CPU against this and wonder about the power draw. Look at what they did to Arrow Lake, I believed most of the slob, bought and then returned, rebought the ultra 7 265K. Then I followed the PCGH tuning guide with 7200MT RAM and am now quite happy with the system, it feels like Raptor Lake with 10% more gaming performance. The only issue is, the moment you return Arrow Lake back to the 66ns memory latency with 115GB/s memory throughput, you land at 270W power draw again. With these settings, the U7 265k competes with a 9950x in MT, ST and is faster in gaming for 285€. Mainboards are extremely cheap as well due to low demand and overproduction? Intel needs to avoid this and immediately deliver a boost profile for those who dare, other people can buy their hardware and lock it down.

 

[511]

If you know Intel, this 150W are the locked down, not tuned, basically constrained specs. Basically a PR move to ward off the techtuber crowd that will bench a 8C CPU against this and wonder about the power draw. Look at what they did to Arrow Lake, I believed most of the slob, bought and then returned, rebought the ultra 7 265K. Then I followed the PCGH tuning guide with 7200MT RAM and am now quite happy with the system, it feels like Raptor Lake with 10% more gaming performance. The only issue is, the moment you return Arrow Lake back to the 66ns memory latency with 115GB/s memory throughput, you land at 270W power draw again. With these settings, the U7 265k competes with a 9950x in MT, ST and is faster in gaming for 285€. Mainboards are extremely cheap as well due to low demand and overproduction? Intel needs to avoid this and immediately deliver a boost profile for those who dare, other people can buy their hardware and lock it down.

TBH thier botched power delivery mechanism for Desktop(DLVR) is to blame for this as well partly

 

[dullard]

I am shocked 52cores are within 150watts of TDP. So let’s add another 100-150 watts and that’s the actual TDP when under full load.

No reason to be shocked at all. Intel already has an 8P + 16E chip with 35 W base power. The 285T. https://www.intel.com/content/www/u...-36m-cache-up-to-5-40-ghz/specifications.html
Double the cores and it could be ~70W base even if nothing else changed.

The only reason Intel CPUs have a reputation for being high power is because reviewers put on massive heat sinks and then test them at high clocks. When run at sane clock speeds, they are not and have not been power hungry. Don't put on a massive heat sink, don't use motherboard settings with unlimited power, don't look solely at peak power, etc. Suddenly they run at quite reasonable power levels.

Arrow lake itself is ~20% more efficient than Raptor Lake. Add in a better node and I see no reason that Nova Lake won't be even more power efficient.

Yes, at full turbo, which only lasts a short time period, Nova Lake will use more than 150 W of power. Your 100 W more estimate is probably pretty accurate for that <1 minute of PL2 time (unless you set the motherboard to use more power and use a heat sink that is capable of dissipating that power, but then, that was your choice to use that much power).

 

[511]

Intel has gotten itself to blame for these things it was their idea for

 

[Thunder 57]

No reason to be shocked at all. Intel already has an 8P + 16E chip with 35 W base power. The 285T. https://www.intel.com/content/www/u...-36m-cache-up-to-5-40-ghz/specifications.html
Double the cores and it could be ~70W base even if nothing else changed.

The only reason Intel CPUs have a reputation for being high power is because reviewers put on massive heat sinks and then test them at high clocks. When run at sane clock speeds, they are not and have not been power hungry. Don't put on a massive heat sink, don't use motherboard settings with unlimited power, don't look solely at peak power, etc. Suddenly they run at quite reasonable power levels.

Arrow lake itself is ~20% more efficient than Raptor Lake. Add in a better node and I see no reason that Nova Lake won't be even more power efficient.

Yes, at full turbo, which only lasts a short time period, Nova Lake will use more than 150 W of power. Your 100 W more estimate is probably pretty accurate for that <1 minute of PL2 time (unless you set the motherboard to use more power and use a heat sink that is capable of dissipating that power, but then, that was your choice to use that much power).

So it's the reviewers fault? Do they review AMD differently, only at strictly stock settings when measuring power? Any credible reviewer includes idle power even though Intel wins that. Surely if they were biased that fact would be omitted?

 

[dullard]

So it's the reviewers fault? Do they review AMD differently, only at strictly stock settings when measuring power? Any credible reviewer includes idle power even though Intel wins that. Surely if they were biased that fact would be omitted?

To paraphrase Epictetus, only fools seek blame. https://www.goodreads.com/quotes/5312902-an-ignorant-person-is-inclined-to-blame-others-for-his

I didn't say there is any fault or any blame at all. Review AMD and Intel with a mid-range heat sink and see what happens. No blame there or bias. Heck, review them with a massive heat sink and unlimited power. Still no blame or bias.

But, to look at a review that includes a massive heat sink and a motherboard set to unlimited power and then assume that the CPU is power hungry for virtually all cases and tout that publicly is where any fact distortion occurs. Heck, just leaving out the fact that the CPUs are not power hungry for many use cases in public posts is the problem.

And it isn't just idle power. Look at actual average power used when you don't use an expensive heat sink and don't use a top end motherboard set to unlimited power. You know, like situation where the majority of computer users are operating.

 

[AMDK11]

Removed

 

[AMDK11]

removed

 

[Fjodor2001]

No reason to be shocked at all. Intel already has an 8P + 16E chip with 35 W base power. The 285T. https://www.intel.com/content/www/u...-36m-cache-up-to-5-40-ghz/specifications.html
Double the cores and it could be ~70W base even if nothing else changed.

The only reason Intel CPUs have a reputation for being high power is because reviewers put on massive heat sinks and then test them at high clocks. When run at sane clock speeds, they are not and have not been power hungry. Don't put on a massive heat sink, don't use motherboard settings with unlimited power, don't look solely at peak power, etc. Suddenly they run at quite reasonable power levels.

Arrow lake itself is ~20% more efficient than Raptor Lake. Add in a better node and I see no reason that Nova Lake won't be even more power efficient.

Yes, at full turbo, which only lasts a short time period, Nova Lake will use more than 150 W of power. Your 100 W more estimate is probably pretty accurate for that <1 minute of PL2 time (unless you set the motherboard to use more power and use a heat sink that is capable of dissipating that power, but then, that was your choice to use that much power).

Exactly, as posted a while ago in this thread from this source:

 

[vanplayer]

There may be a possibility that an 8P/16E die will be manufactured with an Intel 18A.
It may be possible to make a multi-die even with the 18A version of the die.
8p/16e Die is It may be produced by both IFS and TSMC.

18A is meh, just let it be a negligible character in upcoming scene.

 

[511]

18A is meh, just let it be a negligible character in upcoming scene.

source: trust me bro?

 

[Thunder 57]

Exactly, as posted a while ago in this thread from this source:

That's not surprising. That scenario is great for E-cores and that is no secret. From just above in the article:

When encoding in Handbrake, the Core Ultra 9 285K takes a five percent lead over the Ryzen 9 9950X. In this case, however, the AMD CPU runs into its temperature limit due to the completely different type of load (AVX) (the power consumption and temperature measurements using Handbrake as a baseline already showed this). As a result, the processor consumes only an average of 160 watts, while the Core Ultra 9 285K averages 212 watts. The CPU thus gains a five percent advantage here by consuming 32 percent more power—and thus operates significantly less efficiently.

It would seem that is very much an "it depends" scenario. Intel has made great progress here, just look at the 14900k scaling in that graph.

 

[Thunder 57]

To paraphrase Epictetus, only fools seek blame. https://www.goodreads.com/quotes/5312902-an-ignorant-person-is-inclined-to-blame-others-for-his

I didn't say there is any fault or any blame at all. Review AMD and Intel with a mid-range heat sink and see what happens. No blame there or bias. Heck, review them with a massive heat sink and unlimited power. Still no blame or bias.

But, to look at a review that includes a massive heat sink and a motherboard set to unlimited power and then assume that the CPU is power hungry for virtually all cases and tout that publicly is where any fact distortion occurs. Heck, just leaving out the fact that the CPUs are not power hungry for many use cases in public posts is the problem.

And it isn't just idle power. Look at actual average power used when you don't use an expensive heat sink and don't use a top end motherboard set to unlimited power. You know, like situation where the majority of computer users are operating.

I wish we would see more of that. Much more realistic case for most. For most use cases CPU's are plenty powerful today. That's probably why there is such focus at the top end performance. Not saying that's the right way to do it. It just seems that's what their audience expects.