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.

 

[Hitman928]

Yeah, my bad you compared absolute perf (as did WFFC)

I chose 165U / 1360P particularily as they clock very close to the Lunar Lake sample (on average) and are lower TDP parts.

I agree your comparison makes more sense, my post was just in response to the comparison WFFCtech did, which of course used the 185H.

 

[FlameTail]

Will Panther Lake be 18A of 18A-P ?

 

[Magio]

The single core score is quite good and in line with what I expected, the MT score is not as impressive. That's in line with a good 165U result. Probably at lower power as the 165U's that score in that range probably are those with a higher TDP but still, +/-12k is what I was hoping to see. Early results, but let's see.

 

[Gideon]

The single core score is quite good and in line with what I expected, the MT score is not as impressive. That's in line with a good 165U result. Probably at lower power as the 165U's that score in that range probably are those with a higher TDP but still, +/-12k is what I was hoping to see. Early results, but let's see.

IMO considering Lunar Lake TDP and 8c/8t corecount, it's very impressive, it's probably TDP limited. 165U has 12 cores and 14 threads and a 57W power limit for all core loads.

 

[Magio]

IMO considering Lunar Lake TDP and 8c/8t corecount, it's very impressive, it's probably TDP limited. 165U has 12 cores and 14 threads and a 57W power limit for all core loads.

For me I'd say it depends, if it does 10K MT while truly sticking to its 17W PL1, then that's OK if not super impressive, but if it needs to gets close to its 30W PL2 to achieve that then that's a disappointment. A fanless M3 or even M2 can get that score with the same core count.

 

[dttprofessor]

For me I'd say it depends, if it does 10K MT while truly sticking to its 17W PL1, then that's OK if not super impressive, but if it needs to gets close to its 30W PL2 to achieve that then that's a disappointment. A fanless M3 or even M2 can get that score with the same core count.

M1 7700 15W
M2 8700 20W
M3 10395 17W

4 skymont cores is enough for 90% apps, except game.

 

[Magio]

M1 7700 15W
M2 8700 20W
M3 10395 17W

If you check on Geekbench, plenty of M2 Macbook Airs score 10k MT. M3 can go as far as push 12k in a fanless design.

 

[TwistedAndy]

I think the retail Lunar Lake devices will show the results closer to the Intel charts (~14% faster than Meteor Lake on average):

If you check on Geekbench, plenty of M2 Macbook Airs score 10k MT. M3 can go as far as push 12k in a fanless design.

The Geekbench 6 MT test is garbage.

 

[Magio]

The Geekbench 6 MT test is garbage.

I'm not the biggest fan of Geekbench, don't get me wrong. Usually I put a lot more weight on Cinebench, but really no benchmark is perfect.

However right now all we have is this Geekbench score and it's basically a "worst" case scenario in MT, only matching 165U. Hopefully we'll see better scores later and it'll do better in Cinebench.

 

[dttprofessor]

If you check on Geekbench, plenty of M2 Macbook Airs score 10k MT. M3 can go as far as push 12k in a fanless design.

My zenbook14 155H ,12000 40W ,it's not so good, but it can last 6-10hours.
The LNL should be 10K 20-25w, the purpose of intel is lower temperature & 10+ hours even in a 1KG laptop.

 

[coercitiv]

+/-12k is what I was hoping to see.

Here's Alder Lake @ 5Ghz in 8P/8T config getting 12K.

Micro-Star International Co., Ltd. MS-7D25 - Geekbench

Benchmark results for a Micro-Star International Co., Ltd. MS-7D25 with an Intel Core i7-12700K processor.

browser.geekbench.com

For me I'd say it depends, if it does 10K MT while truly sticking to its 17W PL1, then that's OK if not super impressive, but if it needs to gets close to its 30W PL2 to achieve that then that's a disappointment.

Out of curiosity, I ran the test a second time with HWInfo running, watching CPU package power. One MT test went to ~116W (PL2 is 120W), another to ~90W and most others hovered around 60W and few even lower than that. I think I saw one with 45W. From a power perspective Geekbench is not pushing the chips at all.

In fact, it just dawned on me as I typed this, I can test at lower power ceilings. 30W might mess up boosting algo on my 12700K, but 45W is definitely doable and will not affect ST score either.

And the MT score for 8P/8T Golden Cove @ PL1=PL2=45W is.... drumroll... lights moving... 10K!

Micro-Star International Co., Ltd. MS-7D25 - Geekbench

Benchmark results for a Micro-Star International Co., Ltd. MS-7D25 with an Intel Core i7-12700K processor.

browser.geekbench.com

I think it's safe to say LNL does not need 30W for a 10K score. Now excuse me while I go get my desktop config back...

 

[mikk]

I'm not the biggest fan of Geekbench, don't get me wrong. Usually I put a lot more weight on Cinebench, but really no benchmark is perfect.

However right now all we have is this Geekbench score and it's basically a "worst" case scenario in MT, only matching 165U. Hopefully we'll see better scores later and it'll do better in Cinebench.

If 165U matches this core it needs more power. LNL 17W will be 20-35% faster than 165U 15W depending on the MT benchmark. Having only 8 threads won't give exciting MT results but that's not the main thing of LNL. Main thing for most consumers will be battery life/energy consumption in idle and small loads which will help everyone running in battery mode. Also iGPU performance is a big improvement in this TPD range.

 

[DavidC1]

Estimating clocks for the Skymont cluster in Lunar Lake.

SpecInt scales at about 90% relative to clocks. With 38% gain in uarch, at the same power we're at about 3.1GHz and at max performance it's at 3.7GHz for single thread, compared to max 2.5GHz for Crestmont LPE.

For the MT version, Intel claims 2.9x with 2x the cores. Core count scaling is similar for the most part at about 90%, so it leaves 53% for uarch+clocks. That leaves us at 10% higher clocks for Skymont cluster with 2x the core count at the same power as MTL LPE, while having 38%/68% per clock gain. Quite impressive!

Hah, looks like I got the peak clocks for Lunarlake E core correct!

So Lunarlake is where the E core not only covers MT performance but works very efficient at lower load workloads.

 

[Goop_reformed]

So what's the consensus for Lion Cove/Lunar Lake. Is it good or another MTL situation

 

[Thunder 57]

Should be good for its role. That is low power, long battery life, better than expeted performance for its power use.

 

[csbin]

LNL 17W vs 30W

https://twitter.com/x/status/1805718733207298541

 

[poke01]

Wow, Lunar is amazing in gpu performance

 

[Henry swagger]

LNL 17W vs 30W

https://twitter.com/x/status/1805718733207298541

View attachment 101965

Faster than m2 and 5% behide the m3 in single core in geekbench 5.. m3 is 12% faster in multicore.. impressive lion cove looks good 😁

 

[Henry swagger]

https://twitter.com/x/status/1806157105814200820

Battlemage looks excellent.. tsmc nodes are perfect for low power performance.. arrow lake hx will be even better 🙂

 

[poke01]

Amazing what a good node along with good design does huh?

 

[FlameTail]

LNL GPU at 30W matches the X Elite GPU at 30W in Wildlife Extreme. Nice.

Are there Steel Nomad Light numbers?

 

[trivik12]

https://twitter.com/x/status/1806157105814200820

Battlemage looks excellent.. tsmc nodes are perfect for low power performance.. arrow lake hx will be even better 🙂

But Arrow Lake has Alchemist+ and not BMG.

 

[poke01]

But Arrow Lake has Alchemist+ and not BMG.

Why is discrete battlemage coming next year? Or is late Q4? Issues with scaling?

 

[majord]

Must be Cinebench R20?

 

[Magio]

I don't know about all of you, but for me those early scores make it clear why Intel wasn't bragging about benchmarks in their reveal at Computex... We're talking about OK-to-good results at best for now.