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.

 

[poke01]

I tried facebook, reddit, lego.com all with 4x slowdown. No lag 🤷‍♂️

what is the cpu you tested on?

 

[Covfefe]

what is the cpu you tested on?

Stock R5 7600X

 

[DavidC1]

The battery life compared to ARM(Qualcomm) is fully competitive. It's only against Apple that they are behind, because of a much superior OS. Intel chips on Apple had 20-30% better battery life too.

My guess is they are using the Zenbook Duo as a test product because it's a fancy showcase for a new platform.

 

[Joe NYC]

who even said thatt.

It is not really that difficult to figure that out. If you are providing to your customers a majority of their CPUs, they will take seriously all your new CPUs, and they will adopt them.

And adopt them pronto.

You can see it in the server market, where AMD has a large and growing shares, especially at US hyperscalers. When the new CPU (Venice) shows up, it will have broad support right from the start.

Contrast that with AMD weak position in mobile market, and majority of OEMs not even bothering to release a single Strix Halo laptop.

oh ffs it's a new platform in a new swimlane.

It didn't even get a token support of a single model from most of the OEMs, namely Lenovo, Dell, Acer, Microsoft.

If AMD had a strong presence in notebook, if AMD was an indispensable partner, they all would be offering Strix Halo.

this is all worthless since all three are flat to barely higher on 1t perf.
Showing up with nothing ain't worth much.

Intel had Meteor Lake when Strix Halo was released.

Intel improved its competitiveness with these subsequent releases (LNL, PTL, coming up, NVL)

And it is not just the recent past up to now. AMD will be stuck for another (close to) a year with no follow up, no improvement.

It is just a bad business decision by AMD to not invest in this market, where the upside is quite significant. Several billion dollars per quarter upside.

 

[gdansk]

if amd released Zen5 Gorgon point on TSMC N3P they would have beaten Panther lake easily. The cpu is very mediocre

AMD is too lazy. They can't compete in the laptop market. Too much stale IP on a big monolith. Intel is barely stumbling forward yet AMD is already falling behind. They're not a real threat.

The only reason to buy SP over PTL is price and availability. With memory prices dictating end price of laptops more than CPU the only hope AMD has this year is that Intel can't make enough. Which is probably the case.

 

[adroc_thurston]

It is not really that difficult to figure that out. If you are providing to your customers a majority of their CPUs, they will take seriously all your new CPUs, and they will adopt them.

you have no idea what you're talking about.

It didn't even get a token support of a single model from most of the OEMs, namely Lenovo, HP, Acer, Microsoft.

MS is not a real player. Acer is tiny.
AMD key partners are HP/Lenovo/ASUS. All have or are rolling out stxH devices.

You can see it in the server market, where AMD has a large and growing shares, especially at US hyperscalers. When the new CPU (Venice) shows up, it will have broad support right from the start.

completely different markets.

If AMD had a strong presence in notebook, if AMD was an indispensable partner, they all would be offering Strix Halo.

lmao no, it's a novelty product in a new lane.

Intel had Meteor Lake when Strix Halo was released.

Q1'25 was ARL-H, and Raptor had the same 1t anyway.

Intel improved its competitiveness with these subsequent releases (LNL, PTL, coming up, NVL)

No, they're still stuck at 2800-3k 1t.

It is just a bad business decision by AMD to not invest in this market, where the upside is quite significant. Several billion dollars per quarter upside.

you have no idea what you're talking about.
TSM price hikes alone are thinning whatever margin laptop parts had.

AMD is too lazy. They can't compete in the laptop market. Too much stale IP. Intel is barely stumbling forward yet AMD is already falling behind. They're not a real threat.

this will be really funny in a year. oh well.
maybe Intel will move on from 2900 gb6 1t.

 

[gdansk]

this will be really funny in a year. oh well.
maybe Intel will move on from 2900 gb6 1t.

Being competitive every once in a while is the path to Radeon tier market share.
Intel still gets most the interesting designs. They need to be ahead of Intel consistently to get the designs.

 

[adroc_thurston]

Being competitive every once in a while is the path to Radeon tier market share.

It's once every 8Q. Pretty fair.

Intel still gets most the interesting designs. They need to be ahead of Intel for years to get the designs.

they get the designs anyway.
You just deliver the perf/BL and there you go.

 

[Khato]

It is just a bad business decision by AMD to not invest in this market, where the upside is quite significant. Several billion dollars per quarter upside.

While it's bad for consumers I wouldn't claim it's a bad business decision. Prioritizing their AI offerings is the best business decision at the moment as that's where the money is. With a secondary priority of delivering on the next generation console designs. Losing out on client mobile market share for a year is of little consequence.

 

[adroc_thurston]

With a secondary priority of delivering on the next generation console designs.

That's 2027.

Losing out on client mobile market share for a year is of little consequence.

But they're gaining it.
WCL might reclaim msnb share tho.

 

[poke01]

The battery life compared to ARM(Qualcomm) is fully competitive.

there weren't any Qualcomm Elite 2 products tested. When even is that platform launching?

It's only against Apple that they are behind, because of a much superior OS. Intel chips on Apple had 20-30% better battery life too.

Its not a software problem.
Windows laptops esp competent OEMs like Dell actually had BETTER battery life than a Macbook when accounting for Wh and panel differences.

 

[adroc_thurston]

there weren't any Qualcomm Elite 2 products tested. When even is that platform launching?

Glymur doesn't push the BL any further iirc.

 

[mikegg]

I tried facebook, reddit, lego.com all with 4x slowdown. No lag 🤷‍♂️

Chrome DevTools does this: Multiply JavaScript execution time by 4.

It does not slow:

• GPU
• Page rendering
• CSS layout
• DOM painting
• Network
• Input
• OS scheduling
• Browser compositor

So a site like YouTube, FB, or Reddit still:

• Scrolls at 60 or 120 fps
• Renders animations smoothly
• Plays video normally
• Accepts input instantly

Browsers also heavily depend on GPU for rendering nowadays. You’re not going to feel as much lag as if the real Intel SoC is throttled.

 

[DavidC1]

I tried facebook, reddit, lego.com all with 4x slowdown. No lag 🤷‍♂️

This also assumes that throttling is on all PTL, rather than unique to that laptop. Laptop vendors have more than a dozen parameters they can change regarding power management.

Its not a software problem.
Windows laptops esp competent OEMs like Dell actually had BETTER battery life than a Macbook when accounting for Wh and panel differences.

It is a software issue. SteamOS comparisons prove it. Also, Windows 8 is quite a bit lighter compared to 11. I don't know about Icelake, but the previous Intel chips had consistent advantage for Apple. When you consider they add more background and telemetry features every Windows, it makes sense.

Windows laptops you have to wrestle with rogue processes every few days misbehaving that raises CPU package power from 0.5W to 1.5W for no reason. Someone also did a test on Linux vs Windows for a spreadsheet and Windows was several times slower.

 

[Joe NYC]

Why waste time on a shrink when new IP (Zen6-era stuff) is just ahead.

Zen 5 on N3P could (should) have been released at Computex 2025.

Plenty of life left between then and Medusa launch.

If you what to measure that hypothetical timespan, between Zen 5 on N3P to Medusa Halo, it is about 1.5 time of Intel releases.

In other words, Intel releases 1.5 generations in the timespan of what you think is a waste of time.

 

[adroc_thurston]

Zen 5 on N3P could (should) have been released at Computex 2025.

You're wasting people's time porting antiquated core and fabric IP onto N3p. don't.

Plenty of life left between then and Medusa launch.

no, Medusa EEP is late this year.

between Zen 5 on N3P to Medusa Halo, it is about 1.5 time of Intel releases.

Medusa halo isn't a Medusa, it's a Radeon program.

In other words, Intel releases 1.5 generations in the timespan of what you think is a waste of time.

1.5 generations with Raptor Lake level 1t.
wooooooooooooo hoooooooooooo

 

[Joe NYC]

Chrome DevTools does this: Multiply JavaScript execution time by 4.

It does not slow:

• GPU
• Page rendering
• CSS layout
• DOM painting
• Network
• Input
• OS scheduling
• Browser compositor

So a site like YouTube, FB, or Reddit still:

• Scrolls at 60 or 120 fps
• Renders animations smoothly
• Plays video normally
• Accepts input instantly

Browsers also heavily depend on GPU for rendering nowadays. You’re not going to feel as much lag as if the real Intel SoC is throttled.

To test JS bloat, a good web site for that is UK Daily Mail.

The Mrs. "tests" this one quite a bit, leaving 10s of tabs open...

 

[Hulk]

So finally some Panther Lake CPU data.

https://www.tomshardware.com/laptops/gaming-laptops/asus-zenbook-duo-2026-review

Haven't thought about this too deeply yet by this is what I was thinking.

From this Geekbench 6 score and assuming 5.1 for both the 388H and the HX370 we are looking at about +6% throughput per cycle for Couger Lake vs. Zen 5.

4+8+4 also loses by a bit to 12/24 in the Handbrake transcoding test, which would make sense because Handbrake scales very well regardless of codec to about 8 cores. By that point the 388H is deep into Darkmont while the HX375 is still reaching into full P cores, even though the "C" versions are lower clocked.

Battery life looks to be about 11% better.

My quick take on Panther Lake vs. Strix Point? Strix Point is a bit more performant in MT, Panther Lake a bit more with ST. Battery life is better with Panther Lake by +11%. Panther Lake is much stronger (+50%) when it comes to iGPU. If I were buying today I'd make my decision by the rest of the laptop not the CPU. If I could get either CPU in the same laptop I'd buy the less expensive one.

Of course I'm basing this off of some meager data and some healthy assumptions but I bet further testing won't change much in the way of this "napkin" summary.

Assuming Intel can pull another 6-8% IPC out of Nova Lake and AMD gets 12-14% out of Zen 6 it will come to frequency for the ST crown.

 

[Joe NYC]

You're wasting people's time porting antiquated core and fabric IP onto N3p. don't.

Only if you think AMD SoC is already world class, same or better than all the other competitors. In that case, it would be improving perfection.

But in case if you think that AMD still has quite a way to go to equal Mac SoC, then every new iteration is worthwhile.

no, Medusa EEP is late this year.
Medusa halo isn't a Medusa, it's a Radeon program.

Sorry, I meant Medusa 1. I think it should definitely be released this year, preferably mid 2026, 2 years after Strix Point release, but who knows when it actually arrives.

1.5 generations with Raptor Lake level 1t.
wooooooooooooo hoooooooooooo

If you have number of iterations, you can fix / upgrade different parts of the design individually, especially starting from quite an archaic Raptor Lake.

That was to be AMD's huge advantage in Mobile. That Intel was stuck on Raptor Lake, and it was going to be difficult for Intel to un-stuck itself from Raptor.

AMD just did not do enough to take advantage of Intel being stuck on Raptor Lake.

 

[adroc_thurston]

Only if you think AMD SoC is already world class, same or better than all the other competitors. In that case, it would be improving perfection.

But in case if you think that AMD still has quite a way to go to equal Mac SoC, then every new iteration is worthwhile.

You're suggesting no iteration, but a straight shrink. Waste of everyone's time.
Medusa actually has new fabric/CPU/yaddayadda IP.

Sorry, I meant Medusa 1. I think it should definitely be released this year, preferably mid 2026, 2 years after Strix Point release, but who knows when it actually arrives.

It's late'26 EEP; actual volume is 2027.

If you have number of iterations, you can fix / upgrade different parts of the design individually, especially starting from quite an archaic Raptor Lake.

PTL-H has the same-ish 1t perf as the antique RPL-H from 2023.

AMD just did not do enough to take advantage of Intel being stuck on Raptor Lake.

RPL was quite literally the biggest boon Intel had; cheap and nice volume drivers.
Every part after RPL is horribly margin dilutive until *maybe* NVL-U 4+0+4.

 

[Hulk]

You're suggesting no iteration, but a straight shrink. Waste of everyone's time.

Quite standard actually, called a "tick."

 

[adroc_thurston]

Quite standard actually, called a "tick."

Ticks are dead because modern shrink is like 15% speed 30% power best case.

 

[poke01]

, Windows 8 is quite a bit lighter compared to 11.

you mean 10?

View attachment 137370

英特尔Ultra X9深度评测：完美的轻薄本处理器！_哔哩哔哩_bilibili

英特尔最强的核显CPU来啦！快来看看今年轻薄本的新品——Panther Lake-H或者说X9 388H，在大幅增强能效和GPU后，游戏表现能否赶上前几年的独显游戏本呢？, 视频播放量 79010、弹幕量 636、点赞数 4314、投硬币枚数 845、收藏人数 937、转发人数 571, 视频作者 极客湾Geekerwan, 作者简介 爱折腾、超硬核的科技数码频道！商务合作请联系：bd@geekerwan.com，相关视频：Ultra X9 388H性能分析：重铸Intel荣光，Clawdbot硅谷爆火！Mac...

www.bilibili.com

test conditions:

• Test method: background apps logged into QQ, Chrome with Office tabs open
• Browser and QQ kept running
• Same Wi-Fi
• Brightness: 200 nits
• Power mode: “Balanced / Default”

Lunar is better, interesting

 

[Covfefe]

Chrome DevTools does this: Multiply JavaScript execution time by 4.

It does not slow:

• GPU
• Page rendering
• CSS layout
• DOM painting
• Network
• Input
• OS scheduling
• Browser compositor

So a site like YouTube, FB, or Reddit still:

• Scrolls at 60 or 120 fps
• Renders animations smoothly
• Plays video normally
• Accepts input instantly

Browsers also heavily depend on GPU for rendering nowadays. You’re not going to feel as much lag as if the real Intel SoC is throttled.

I halved my CPU's clockspeed in BIOS, and ran the test again. The LCP took 780ms, which is only took 7% longer than stock. So it looks like chrome devtools CPU throttling is inaccurate, just not in the direction you expected.

For GPU rendering, input, OS scheduling, etc. I don't see your point. We are discussing Intel Laptop CPUs scoring lower in javascript benchmarks. They might also throttle GPU or have bad input latency, but I've seen no reputable claims for that.

 

[Hulk]

you mean 10?

View attachment 137390
test conditions:

• Test method: background apps logged into QQ, Chrome with Office tabs open
• Browser and QQ kept running
• Same Wi-Fi
• Brightness: 200 nits
• Power mode: “Balanced / Default”

Lunar is better, interesting

True but a massive improvement from ARL.

But PTL is 18A vs N3B for LL compute tile. If everything but the LPE was powered down during this test then I would expect at least equal power efficiency to Lunar Lake if 18A is better than N3B?

But then again LL has on package memory...