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.

 

[Thunder 57]

Intel would have to lose for many many years for AMD to get to 50% market share then.

Intel seems to be doing enough damage on its own to make that happen. Their 18A better be as promised. The other threat is ARM. I like x86 because of its backwards compatibility. Also the fact that it isn't a "walled garden".

 

[poke01]

Intel seems to be doing enough damage on its own to make that happen. Their 18A better be as promised. The other threat is ARM. I like x86 because of its backwards compatibility. Also the fact that it isn't a "walled garden".

A bit off topic but ARM isn't a walled garden. You can get ARM worksations from System76 with upgradeable CPUs/GPUs/RAM which run Linux. Its just no one cares about Windows on ARM because Qualcomm had an exclusivity period with MS. Apple's idea of ARM computer is different from others like System76.

In the future who knows a RISC-V/ARM core might be so powerful that even with a 20% emulation hit its still faster than then contemporary x86 designs.

Thelio Astra arm64 - Linux Laptops, Desktops, and Servers

system76.com

 

[CakeMonster]

I can understand Intel being in a challenging situation with their initial product on a re-design, and waiting for their own next node and higher prices at TSMC. But they better be able to iterate on this quite quickly and steadily to keep gamers interested.

 

[Josh128]

Techspot uses HUB results I think

Indeed, for some reason my eyes saw "Techpowerup". 😂

 

[Saylick]

May God have mercy on Pat's soul...

https://twitter.com/x/status/1850039772498239821

https://twitter.com/x/status/1850040480370917744

 

[alcoholbob]

May God have mercy on Pat's soul...

https://twitter.com/x/status/1850039772498239821

https://twitter.com/x/status/1850040480370917744

It's only a problem if there's actually a lot of demand for the 285K...

 

[Saylick]

It's only a problem if there's actually a lot of demand for the 285K...

Ouch.

 

[gdansk]

265K is the more desirable part anyway isn't it?
And should be easier to bin.

 

[RTX2080]

May God have mercy on Pat's soul...

https://twitter.com/x/status/1850039772498239821

https://twitter.com/x/status/1850040480370917744

It's only a problem if there's actually a lot of demand for the 285K...

Not speaking demand, but 285k seems having supply problem which 265k/245k don't have. It may reflect the cost issue of ARL though.

 

[TESKATLIPOKA]

Not speaking demand, but 285k seems having supply problem which 265k/245k don't have. It may reflect the cost issue of ARL though.

265K and 245K are the same chip as 285K If I am not mistaken.
If there is a supply issue, then only because there are many faulty chips.

 

[adroc_thurston]

265K is the more desirable part anyway isn't it?

Not for Intel lmao.

265K and 245K are the same chip as 285K If I am not mistaken.

Different yield parameters. 5.7 on a foundry node consistently is hella achievement.

 

[TESKATLIPOKA]

Not for Intel lmao.

Of course not If It's $195 cheaper, but I as a customer don't care about that.

 

[Kocicak]

I tested a power limited Arrow lake here:

Testing ARROW lake - power limited

I got myself a testing PC with 265K, Strix E board and ordinary RAM (32GB, 6000 MHz, CL36) and tested the CPU with two power limits 100 and 160W, and compared it to 13900KS with 4 E cores disabled, both with HT on and off. As you can see, the results are pretty consistent, both apps in both...

www.techpowerup.com

 

[lopri]

1.480 Vcore is in my opinion too much for Intel 7, much less a smaller node, under ambient for day-to-day use. You are practically begging for degradation. Everybody has a different opinion but I keep my 14900K under 1.3V and it's generally under 1.2V when under load.

This person got a great sample and is pushing it to the limit. It's great to show what ARL can do but not really useful for people that want to use these CPU's for longer than a few weeks.

From what I have seen online 1.4~1.5V (transient) is not an outlier. I think Ryzens do it, too.

But what I cannot get my head around is how 1.5V/250W is only reporting 80C max temp.

 

[inf64]

It looks like intel will be in the rear view mirror even after Panther Lake, especially in gaming. It's sad since this means AMD will be commanding the prices for their X3D parts which will be priced higher than most of us would like to see. It was a solid opportunity for intel after flopped Zen 5 launch, but they managed to out-flop AMD with this one.

 

[SiliconFly]

.

 

[maddie]

But what I cannot get my head around is how 1.5V/250W is only reporting 80C max temp.

Removing that much power from the available surface area should mean higher temps. A worthy investigation for someone.

 

[mzocyteae]

A bit off topic but ARM isn't a walled garden. You can get ARM worksations from System76 with upgradeable CPUs/GPUs/RAM which run Linux. Its just no one cares about Windows on ARM because Qualcomm had an exclusivity period with MS. Apple's idea of ARM computer is different from others like System76.

In the future who knows a RISC-V/ARM core might be so powerful that even with a 20% emulation hit its still faster than then contemporary x86 designs.

Thelio Astra arm64 - Linux Laptops, Desktops, and Servers

system76.com

Surely AmpereOne is not a walled garden, but it pretty much isn't a valid choice as a desktop/workstation:
1. Low availability - few MB choice and retailers despite the cpus are launched long ago.
2. High cost - list price is way too expensive when comparing to msdt; total system cost won't be much better than epyc.
3. Questionable performance: lack of review (and MT/power is not promising); doomed ST; lack of cpu 2017 submission.
4. win32/linux binary support: some oss projects exists, but in they definitely can't do as well as wine or vmware workstation.
Maybe 1 & 3 will get better in the future, but 2 & 4 is unlikely.

 

[majord]

It looks like intel will be in the rear view mirror even after Panther Lake, especially in gaming. It's sad since this means AMD will be commanding the prices for their X3D parts which will be priced higher than most of us would like to see. It was a solid opportunity for intel after flopped Zen 5 launch, but they managed to out-flop AMD with this one.

but I thought AMD were " never again to be in the windshield" :

 

[SiliconFly]

.

 

[poke01]

I tested a power limited Arrow lake here:

Testing ARROW lake - power limited

I got myself a testing PC with 265K, Strix E board and ordinary RAM (32GB, 6000 MHz, CL36) and tested the CPU with two power limits 100 and 160W, and compared it to 13900KS with 4 E cores disabled, both with HT on and off. As you can see, the results are pretty consistent, both apps in both...

www.techpowerup.com

whats more interesting, is your username over at TPU.

I love it lol

 

[DAPUNISHER]

265K is the more desirable part anyway isn't it?
And should be easier to bin.

I do not think that word means what you think it means. In retail it is going to collect dust with the Zen 5 parts. All these latest generations from the IHVs have done is help sell the older SKUs. Going to take a significant price drop to move them. It would also be a BIG help if Intel would confirm 1851 isn't one and done.

 

[poke01]

ì ë¡ì°ë ì´í¬ì ê´ë ¨í´ì David Huangìê² ë¬¼ì´ë´¤ìµëë¤.

Q:ì§ë¬¸ì´ ììµëë¤. í ë¬ ì ì ê²ìí 루ëë ì´í¬ ì±ë¥ 리뷰ìì, ëì¼ í´ë ìëì ë ëì°ë ì½ë¸ë¥¼ ì¬ì©íâ¦

quasarzone.com

Interview with David Huang.

these answers are interesting:

Q: I have a question. In the Luna Lake performance review posted a month ago, the Luna Lake 258V was about 12% faster than the Meteo Lake using the same clock speed Redwood Cove. But why did Arrow Lake only improve by about 2% compared to the previous work? Shouldn't Arrow Lake show better performance than Luna Lake in terms of IPC? Can you explain the reason to me?

A: Luna Lake's L3 latency is much better than this.

Q: In the case of Luna Lake, is IPC better because the L3 latency is much lower than Arrow Lake? However, I personally don't understand that the IPC improvement is only 2% compared to the previous work when considering the expansion of the interior of Arrow Lake's architecture.

A: The L3 latency of Luna Lake and Rapter Lake is about 55 cycles, while Arrow Lake (265K) is about 73 cycles and Meteo Lake (155H) is 80 cycles.

Q: I wonder how the result came out with such a low result. And why does Arrow Lake perform well on benchmarks like Cinebench R23/R24 or Geekbench?

A:Cinebench and Geekbench are mostly executed within the L2 cache. SPEC is generally much more sensitive to cache and memory.

 

[tsamolotoff]

is this even explicable? )

There can be wildly different results depending on the test used, i.e. there is a popular canned benchmark in the steam workshop that many reviewers use which shows high FPS for almost any CPU and is basically a gpu benchmark while others (i2hard for example) use real-world demo files that were recorded during WGL championships or other high-ranking events where _sometimes_ CPU is more of a limit. Also, CS2 unlike CSGO has some particle effects processed via GPU now so the FPS lows might be actually caused by GPU, not CPU now. Also, if you use several monitors DWM sometimes breaks down and this tanks your FPS considerably as Windows attempts to sync game with DWM which leads to characteristic 'saw line' frametime which can easily noticed when you do tests manually, but it naturally would pass unnoticed while using automated benchmark tools.

 

[tsamolotoff]

HW Unboxed actually discusses this phenomenon here: in some more GPU-heavy game segments ARL looks nice, but when the game starts to put more pressure on the CPU, situation is drastically reversed.