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.

 

[OneEng2]

N3B was bad anyway, we've seen a tweet of TSMC comparing the evolution of various nodes from N7 to A14 in terms of performance and efficiency, and they use N3E instead of N3B for a reason.

N3B was very good. In fact, from a performance standpoint, it was the most performant variant TSMC created.

N3B was one of the nodes that tsmc didn't do well and was late it is still better than N5 by double digit though and has decent density but the cost was not good.

Yep. It was too expensive.

 

[poke01]

In fact, from a performance standpoint, it was the most performant variant TSMC created.

nope its N3X or for actual volume node its N3P.

 

[mikegg]

Qualcomm ain’t getting any meaningful market share with those pathetic iGPUs.

No Nvidia option or any big iGPU. Why would any sane person buy a 18 core cpu and pair it with a weak iGPU?

They still don’t get it. Panther lake despite its okay cpu is still a better buy thanks to its iGPU.

Nah, iGPUs are way less important than you think for the mass market. It's only on technical internet forums that nerds care.

 

[511]

N3B was very good. In fact, from a performance standpoint, it was the most performant variant TSMC created.

It is the worst N3 node tbh for a reason Costly not much perf/watt improvement vs N4P the only thing going for it was density

 

[511]

Nah, iGPUs are way less important than you think for the mass market. It's only on technical internet forums that nerds care.

Same for Geekbench

 

[okoroezenwa]

Same for Geekbench

Just Geekbench? Because it applies to benchmarks in general.

 

[mikegg]

Same for Geekbench

Good thing Arm computers actually feel faster and way more efficient in real life.

Use an AMD/Intel laptop for a day. Then a Macbook. You'll never go back.

Qualcomm stands a way better chance of duplicating the Macbook experience than Intel and AMD could ever dream of.

 

[511]

Good thing Arm computers actually feel faster and way more efficient in real life.

Have you tried LNL or are you still stuck at RPL

Use an AMD/Intel laptop for a day. Then a Macbook. You'll never go back.

I have both and I won't use Mac Cause for personal use ever I can't use half of my stuff and tools.

Qualcomm stands a way better chance of duplicating the Macbook experience than Intel and AMD could ever dream of.

Yeah that's not true only reason you get MacBook experience is Apple actually integrating stuff correctly which Intel/AMD can't do and neither Qcom

 

[poke01]

I have both and I won't use Mac Cause for personal use ever I can't use half of my stuff and tools.

Yeah for people who only main laptops and use win x86 apps, macbooks aren't gonna cut it.

integrating stuff correctly which Intel/AMD can't do and neither Qcom

even with Linux you won't get the same battery life as you do on Windows with Lunar lake.

 

[511]

even with Linux you won't get the same battery life as you do on Windows with Lunar lake.

i was talking about system level integration like every hardware component

 

[poke01]

i was talking about system level integration like every hardware component

x86 can do it too. PC OEMs don’t give a crap.

 

[adroc_thurston]

PC OEMs don’t give a crap.

Oh but they do.
No one likes paying platform premiums, though.

 

[Nothingness]

It's 30%.
10% is the CCD freebie you can get now with something like 9955HX.

Trying to understand. On what ST benchmark is 9955HX already 10% faster? And faster than what? I think that 9950x3d should be the basis, but might be wrong.

Even not taking into account these 10%, 30% is a lot. As for the previous 40% previous claim for Zen5, I'd be happy that AMD delivers that increase, but will wait for evidence at release.

 

[Kepler_L2]

Trying to understand. On what ST benchmark is 9955HX already 10% faster? And faster than what? I think that 9950x3d should be the basis, but might be wrong.

Even not taking into account these 10%, 30% is a lot. As for the previous 40% previous claim for Zen5, I'd be happy that AMD delivers that increase, but will wait for evidence at release.

Geekbench https://browser.geekbench.com/v6/cpu/compare/15259618?baseline=15264690

 

[adroc_thurston]

Trying to understand. On what ST benchmark is 9955HX already 10% faster?

I think "GB6" stands for GB6 pts.
The average good run STX/KRK1 is 2850-ish.
Fire Range is 3150-3200.
Basically you can tape a CCD to STX1 for a 10% freebie which is *exactly* what they're doing for MDS1 hi.

30% is a lot

You can get there with two shrinks alone.
Again, QC N4p -> N3X 3-2 FF is 19% freq with very very mild Vmax cranking.

As for the previous 40% previous claim for Zen5,

It did that and much much more in server.

 

[511]

Oh but they do.
No one likes paying platform premiums, though.

they back out many time cause it will increase BoM

 

[mikegg]

Have you tried LNL or are you still stuck at RPL

Yes. Extreme throttling when unplugged to save battery life. Poor ST and very poor MT performance. Expensive for what you get.

 

[adroc_thurston]

Yes. Extreme throttling when unplugged to save battery life. Poor ST and very poor MT performance. Expensive for what you get.

Bait used to be believable

 

[Nothingness]

Geekbench https://browser.geekbench.com/v6/cpu/compare/15259618?baseline=15264690

Was the comparison against Strix Point as a basis? I missed that.

Anyway enough of off-topic stuff, this belongs to Zen 6 thread

 

[511]

Yes. Extreme throttling when unplugged to save battery life. Poor ST and very poor MT performance. Expensive for what you get.

Proof for ST The MT Performance is not great yes

 

[Joe NYC]

Was the comparison against Strix Point as a basis? I missed that.

Anyway enough of off-topic stuff, this belongs to Zen 6 thread

Actually, not so much off topic, since some of the Zen 6 Medusa configurations will get the optional CCD.

 

[adroc_thurston]

Actually, not so much off topic, since some of the Zen 6 Medusa configurations will get the optional CCD.

Offtopic relative to this thread.

 

[poke01]

Yes. Extreme throttling when unplugged to save battery life. Poor ST and very poor MT performance. Expensive for what you get.

Lunar lake doesn’t do that that much, if at all. The 256v is more than okay for normal cases. Is it M5 level? No and it won’t be till they get rid of the Cove P core.

 

[OneEng2]

nope its N3X or for actual volume node its N3P.

Neither of which was available for ARL/NVL. It was more performant than N3E.

It is the worst N3 node tbh for a reason Costly not much perf/watt improvement vs N4P the only thing going for it was density

Worst how (other than cost as discussed)?

 

[Hitman928]

Neither of which was available for ARL/NVL. It was more performant than N3E.

It was not, just more dense.