Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads

[Tigerick]

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.





Intel Core Ultra 100 - Meteor Lake



As mentioned by Tomshardware, TSMC will manufacture the I/O, SoC, and GPU tiles. That means Intel will manufacture only the CPU and Foveros tiles. (Notably, Intel calls the I/O tile an 'I/O Expander,' hence the IOE moniker.)

 

[SiliconFly]

.

 

[desrever]

They should at least make sure their new CPU works with the latest windows version. ARL can't even run on 24H2 without crashing as mentioned by multiple reviewers.

 

[Hitman928]

They should work with M$ more closely. PC World says one simple setting fixes some important issues. i.e, change the power mode from "Balanced" mode to "Best Performance" mode.

Warning! Most Windows power settings slow Intel's new CPUs to a crawl

Intel's latest Core Ultra processor, Arrow Lake, suffers from a severe performance drop when run in anything but the "High Performance" mode within Windows 11.

www.pcworld.com

Gamer's Nexus, HWUB, and I believe TPU were all aware of the power plan issue and made sure that they didn't test with Balanced mode on 24H2, so it's not effecting their results.

 

[Hitman928]

Yep, TPU's results weren't effected. W1zzard shared a graph in their forum about what it looks like on 24H2 on balanced mode (red bar) compared to their review results (green bar). I'd say this was not an issue in most reviews, or at least not in any that I've seen, as otherwise we'd be talking about ARL struggling to match Zen 2. Hopefully those who buy ARL figure out quickly why their performance is so bad and switch the power plan. Reviewers are doing potential ADL customers a disservice by not highlighting this more.

 

[moinmoin]

Still Arrow Lake is poorly received due to poor gaming performance and a bunch of regressions here and there, some of which seem latency-related and others which might be scheduling problems.

If that's the case there would be nothing to be 'fix' in the future just like Meteorlake.

Both true, I think everybody got that Raptor Lake to Arrow Lake is generally considered a regression. My question is whether there really are no package/uncore improvements between Meteorlake and Arrow Lake.

After all Intel itself considered Meteorlake too bad for a desktop launch. But Arrow Lake with (slightly) improved P- and (vastly) improved E-cores as well as removed SMT while presumably keeping package/uncore unchanged makes it fine for a desktop launch? I honestly can't imagine package/uncore being completely unchanged between Meteorlake and Arrow Lake, surely there have to be some improvements?

 

[gaav87]

Bit off topic but i found this shipping manifest. Nvidia is shipping gddr7 16G from micron last month will they sit on them for 4+ months or launch is sooner then expected ? Idk whats the policy for buying memory before launch is it usually long before ?

 

[Timorous]

Nope it is a disaster a Mobile Product forced on Desktops

The last time Intel pushed a mobile base design to desktop we ended up with Conroe. This is a far cry from that.

 

[PJVol]

Re: reviews...
is this even explicable? )

 

[SolidQ]

is this even explicable? )

China love intel, so their results always weird

 

[desrever]

China love intel, so their results always weird

I wouldn't ever trust any Chinese reviews of anything tbh.

 

[Saylick]

I wouldn't ever trust any Chinese reviews of anything tbh.

Not even Geekerwan's stuff?

 

[DrMrLordX]

AMD managed to tread water this year with their cpu's and STILL managed to beat Intel. That's just sad.

AMD didn't tread water. Ponder that Zen5 was really created for Turin, and then consider that the gamer version of the chip for us plebs hasn't even hit the market yet.

Hopefully this will turn out like Zen1 and lead to significantly better products in the future.

Not possible. Zen1 was 40-50% more performant than its predecessor, catapulting a dying company into the spotlight. Arrow Lake is no such thing.

Zen 5 and Arrow Lake are so remarkably similar in performance, price, power consumption, and disastrous / embarrassing launches

See above. The idea that Zen5's launch is a disaster is a joke. Don't try to make Arrow Lake look good by gaslighting people, thanks.

Also can we all remember that this isn't a "versus" thread? Arrow Lake is (so far) not doing well even compared to previous-gen Intel products, no need for anyone to pretend some other company's launch is showing an industry-wide trend.

 

[Thunder 57]

There is the implied cavaet that the product actually works and doesn't blow up. Anand was never wrong!

True enough. I have a 1TB 7200.11 that should be long dead, yet it still works after a decade. I don't trust any data with it though.

 

[Hitman928]

Re: reviews...
is this even explicable? )

View attachment 110311

Is it possible? Yes. Is it an honest and fair comparison? Doubtful.

 

[coercitiv]

True enough. I have a 1TB 7200.11 that should be long dead, yet it still works after a decade. I don't trust any data with it though.

So what do you keep on it, early Arrow Lake reviews?

 

[ondma]

Also with all the layoffs and project cancellations at Intel, it makes me shudder to think the next socket after Arrow Lake might be even worse relative to next-gen AMD offerings. Scary to think Alder Lake might be the high point of Intel for a very long time.

yea, I said a long time ago. Intel desperately needs a home run at some point to regain leadership. They had a golden opportunity with ARL, since Zen 5 was underwhelming at best. Instead, Intel screwed the pooch with an unbelievably poor product. I would even have not been greatly disappointed had ARL been comparable or only a few percent ahead of RL if it had been very efficient and rock solid stable. If you only want to run Cinebench, it is a fine product. It is good in some other apps, bad in others, and an inconsistent disaster in gaming. Even worse, with Intel cutting back on so many resources, I question if they will ever recover. I know a lot of people dont like intel, but you can see what lack of competition is doing already with rising prices on AMD X3D processors.

This release reminds me of the old Woody Allen quote. To paraphrase: "Arrow Lake is divided into the horrible and the miserable".
Gaming performance is "horrible" and the rest of the performance is "miserable". The only consistent gain is in power usage, and even then with a huge node jump and a redesigned core, it still uses too much power.

 

[poke01]

Look at the HUB video right below where you posted this.

Techspot uses HUB results I think

 

[Hitman928]

Techspot uses HUB results I think

Correct. Techspot is the written version of HWUB.

 

[poke01]

If Lion Cove IPC is so high above ZEN5, why doesn't it show in application performance

It’s does only well in one application, Cinebench 2024 and nothing else lol

 

[poke01]

which is making me think that Arrow Lake's good showing in CB24 is more about its bandwidth than its processing prowess while CB23 seems to be more processor constrained.

CB2024 is also processor intensive but also takes memory bandwidth into account.

 

[poke01]

Not even Geekerwan's stuff?

IMO, Geekerwan is good but for their SPEC tests they need to list compiler flags and versions. On the whole, they are great and I trust them.

 

[poke01]

https://twitter.com/x/status/1849838602920002007

Ouch. Translation: I have nothing to say... It's basically like Intel used 3nm to make Zen4.

 

[Thunder 57]

So what do you keep on it, early Arrow Lake reviews?

I make sure anything important is backed up, including that drive. It holds mostly old games and DVD rips. It has quite a history though with 37674 powered on hours (over 4 years). And spun it up over 17,000 times. What a trooper!

Spoiler: HDD

yea, I said a long time ago. Intel desperately needs a home run at some point to regain leadership. They had a golden opportunity with ARL, since Zen 5 was underwhelming at best. Instead, Intel screwed the pooch with an unbelievably poor product. I would even have not been greatly disappointed had ARL been comparable or only a few percent ahead of RL if it had been very efficient and rock solid stable. If you only want to run Cinebench, it is a fine product. It is good in some other apps, bad in others, and an inconsistent disaster in gaming. Even worse, with Intel cutting back on so many resources, I question if they will ever recover. I know a lot of people dont like intel, but you can see what lack of competition is doing already with rising prices on AMD X3D processors.

This release reminds me of the old Woody Allen quote. To paraphrase: "Arrow Lake is divided into the horrible and the miserable".
Gaming performance is "horrible" and the rest of the performance is "miserable". The only consistent gain is in power usage, and even then with a huge node jump and a redesigned core, it still uses too much power.

It may seem that way, and there are fans out there. I think what most reasonable people want is for Intel and AMD to be competetive with a near 50/50 market share. I think most don't want to have a Nvidia situation with CPU's. We've seen how that works out with GPU's.

 

[Saylick]

It may seem that way, and there are fans out there. I think what most reasonable people want is for Intel and AMD to be competitive with a near 50/50 market share. I think most don't want to have a Nvidia situation with CPU's. We've seen how that works out with GPU's.

Yes, this is exactly what I want. It’s a super idealistic desire which will never happen, but two companies (or more) with equal talent, resources, and profit competing for our dollars is best for the consumer. One, two, even three generations of an underdog winning over an incumbent is nowhere near what it takes to tip the scales back to equal footing.

 

[desrever]

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