24-Core Intel Core Ultra 9 285 Falls Short of 8-Core Ryzen 7 9700X in Geekbench Leak

[rv8000]

Just wait for the reviews people…

 

[GhostRyder]

Well it is the lower SKU and while I am not the biggest fan of this benchmark it probably has more to do with the fact the chip is being held back for TDP numbers compared to the much higher performing K SKU. That being said, I will be curious this generation because they are supposed to have some decent improvements that don't require pushing the P cores to high clocks to make up the single threaded performance.

 

[phints]

Interested in 265K for sure, no interest in these more expensive higher tiers that won't benefit much if anything. But looks like AMD is still the efficiency king even with all the claims from Intel to the contrary.

Exciting times for CPUs though looking forward for reviews. They better all be tested on 24H2 not some older OS.

 

[Neo_Morpheus]

If the cpu was an AMD, the responses would definitely be a lot less…civil and instead would include adjectives like “typical AMD garbage “, “DOA” by the way, thats Ngreedia Unboxed favorite when reviewing an AMD product, etc.

That said, i agree with the point of waiting for proper and hopefully, unbiased reviews.

 

[JIWIL]

Yeah, this is just so far off the general ballpark it shouldn't even be news. Like some rando journalist ran all the benches simultaneously while out for a cuppa.

 

[maxfly]

If the cpu was an AMD, the responses would definitely be a lot less…civil and instead would include adjectives like “typical AMD garbage “

Wrong. That's nothing more than a weak ass attempt at trying to start a flame war. TPU doesn't consist of a bunch of fanboys waving flags. We're better than that. If a benchmark is garbage, it's called out for what it is. The CPU makes no difference.

 

[persondb]

Look how Lunar Lake ended up.

IDK what you are talking about, Lunar Lake is pretty good for what it is.

 

[yfn_ratchet]

Meh, Geekbench doesn't say much unless you're comparing like-for-like, in which case you can just use any other benchmark where results can be more steadily compared. If this kind of disparity was appearing in Cinebench, Y-Cruncher, or software encode I'd be raising an eyebrow and wondering if desktop Arrow Lake is trying to pull a Ryzen 9000 again.

 

[wolf]

Wrong. That's nothing more than a weak ass attempt at trying to start a flame war. TPU doesn't consist of a bunch of fanboys waving flags. We're better than that. If a benchmark is garbage, it's called out for what it is. The CPU makes no difference.

Preach

 

[Minus Infinity]

LOL yesterday I say headline showing 265K beating 9950X in R24. Can we stop will all the BS rumours and just wait for the release.

 

[Neo_Morpheus]

LOL yesterday I say headline showing 265K beating 9950X in R24. Can we stop will all the BS rumours and just wait for the release.

It brings the boys out to the yard and helps me live rent free on some of their heads.
But yes, these kind of articles shouldn’t be posted without more data or proof.

 

[LittleBro]

IDK what you are talking about, Lunar Lake is pretty good for what it is.

Well, Intel promised at least 50% iGPU performance improvement over Meteor Lake iGPU (Core Ultra Series 100, Intel Arc 8 iGPU) last year. Later, there was newer slide with iGPU performance being only 30% better. Also, there were some bold statements made regarding how ARM (Apple and Qualcomm) will be beaten in terms of batter life of notebooks. Results here. Single core performance is okay, but multi core performance sucks. It could have been much better with HT enabled.

Even bolder statement was made towards Lunar Lake iGPU being the most powerful iGPU in the industry. However, Strix Point's iGPU wins in most of modern games:

(Tested at full speed mode, e.g. 28W TDP.)

https://www.notebookcheck.net/Intel-Arc-8-Cores-iGPU-Benchmarks-and-Specs.782930.0.html

https://www.notebookcheck.net/Intel-Arc-Graphics-140V-Benchmarks-and-Specs.854991.0.html

Intel Lunar Lake iGPU analysis - Arc Graphics 140V is faster and more efficient than Radeon 890M

An analysis by Notebookcheck of the Lunar Lake Core Ultra 7 258V coupled with the new Intel Arc Graphics 140V iGPU.

www.notebookcheck.net

The thing is, Intel had been presenting this Lunar Lake as a sort of "everything-beating-and-with-battery-ever-lasting" premium laptop solution.

Boottom line:
Multicore performance is weak. iGPU average performance improvement over Arc 8 (in productivity tasks) is just around+13%, gaming performance on average is better by 26% over Arc 8 according the video above, by NotebookCheck it's better on average by only 10% over Arc 8. Previous bold promises were at first +50% over Arc 8, later +30% over Arc 8. Lunar Lake delivers great single core performance. Battery life and iGPU efficiency is a huge step forward compared to Intel Ultra 100H CPUs/iGPUs and AMD Strix Point. All in all, it's a mediocre product: promised iGPU performance not delivered, ARM battery life not beaten, AMD Strix Point not beaten in most games (tested at 28W = full speed for Lunar Lake).

 

[kawice]

Where are the real world tests to back up your claim? Or are you using your crystal ball?

The single CCX designs have been dominating the gaming charts, whilst using a fraction of the power. Have you not seen?

You mean the few game developers that actually care for gamers and optimize their game for all platforms? Or a few AMD sponsored games that obviously aren't optimized for competition?
To name a few even Cyberpunk 2077 ran like a$$ on AMD shortly after launch. That's how good AMD is.

90% of games are optimized for Intel and don't run well on multiple CCX. The single CCX you mention was like one midrange CPU from AMD's line in their previous gen. WoOw. Looking for more your brilliant thoughts. Please do share.

 

[SL2]

That's sassy, girl!

Let me tell you how it works: The 7800X3D have one single CCX, and not really mid range in terms of gaming performance.

Try google before posting next time lol

Say hi to 2019 from me!

To name a few even Cyberpunk 2077 ran like a$$ on AMD shortly after launch.

Oh you mean the launch FOUR YEARS ago? I hope you're feeling better now. Maybe time to connect to internet and get those updates.

Runs like ass my ass.

 

[ThomasK]

That's sassy, girl!

Say hi to 2019 from me!

Beware of contradicting late 2000's Gen Z on the internet with FACTS, they usually get angry and whine on TikTok.

 

[LittleBro]

90% of games are optimized for Intel and don't run well on multiple CCX. The single CCX you mention was like one midrange CPU from AMD's line in their previous gen. WoOw. Looking for more your brilliant thoughts. Please do share.

Please, don't mistake "Let's roll" for "Let's troll".

Games don't give a f*** about modular design of a processor. They care about cores, they don't know which core will OS assign to them. That's CPU and/or OS scheduler's thing. It's ill-advised to optimize a game for particular MCM design. When you take into account a fact, that a chiplet will hold different amount of cores in the future, that would basically mean a never ending optimization nightmare for game developers.

Intel have been using chiplet design for almost 2 years now, started with Meteor Lake and Sapphire Rapids. On the contrary to monolithic approach, chiplets enable manufacturer to reuse partially defective clusters (CCX or tiles) for other purposes. Monolithic chip production was going well with small (up to 4 cores) CPUs, but as CPU size raises, so raises the production defect rate.

Yes, there comes a latency penalty with chiplet approach, but no one will care anymore. Chiplets generate more money than monolithics and money makes the world go round, not inter-chiplet latencies.

 

[SL2]

Beware of contradicting late 2000's Gen Z on the internet with FACTS, they usually get angry and whine on TikTok.

Nothing will stop me when I see that level of ignorance and attitude.

Please, don't mistake "Let's roll" for "Let's troll".

But, but, he said something about a few games, then continued to mention ONE game. At launch. Four years ago.

https://www.notebookcheck.net/Intel-Arc-8-Cores-iGPU-Benchmarks-and-Specs.782930.0.html

https://www.notebookcheck.net/Intel-Arc-Graphics-140V-Benchmarks-and-Specs.854991.0.html

Intel Lunar Lake iGPU analysis - Arc Graphics 140V is faster and more efficient than Radeon 890M

An analysis by Notebookcheck of the Lunar Lake Core Ultra 7 258V coupled with the new Intel Arc Graphics 140V iGPU.

www.notebookcheck.net

The thing is, Intel had been presenting this Lunar Lake as a sort of "everything-beating-and-with-battery-ever-lasting" premium laptop solution.

Boottom line:
Multicore performance is weak. iGPU average performance improvement over Arc 8 (in productivity tasks) is just around+13%, gaming performance on average is better by 26% over Arc 8 according the video above, by NotebookCheck it's better on average by only 10% over Arc 8. Previous bold promises were at first +50% over Arc 8, later +30% over Arc 8. Lunar Lake delivers great single core performance. Battery life and iGPU efficiency is a huge step forward compared to Intel Ultra 100H CPUs/iGPUs and AMD Strix Point. All in all, it's a mediocre product: promised iGPU performance not delivered, ARM battery life not beaten, AMD Strix Point not beaten in most games (tested at 28W = full speed for Lunar Lake).

I've seen those NBC reviews earlier, but didn't take a good look at it until now.

I dunno, but mobile CPU reviews are too heavily constained by individual laptop SKU characteristics. It's way too early to make a conclusion.

Just look at last generation. Now I'm not saying this is an apples to apples comparison, these three have an 8 core IGP, with two running at 2250 MHz (155h, 28W), and the other one is 2350 MHz (185H, 45W), different RAM speed, and last but not least, using different laptop chassis with different cooling and firmware. Also, I have no idea if they have the same power limitations.

That's not important, all I'm saying is that one single SKU can't tell us everything. Despite the minimal differences, one is 53% faster than the other, one is 29% faster. Yes, 45W is quite a lot more than 28W, I know. Keep focused on the two 155H models instead.

Chances are that something faster than the S 14 exists, just like the Magicbook is so much faster than the Galaxybook.

 

[LittleBro]

Nothing will stop me when I see that level of ignorance and attitude.


But, but, he said something about a few games, then continued to mention ONE game. At launch. Four years ago.


I've seen those NBC reviews earlier, but didn't take a good look at it until now.

I dunno, but mobile CPU reviews are too heavily constained by individual laptop SKU characteristics. It's way too early to make a conclusion.

Just look at last generation. Now I'm not saying this is an apples to apples comparison, these three have an 8 core IGP, with two running at 2250 MHz (155h, 28W), and the other one is 2350 MHz (185H, 45W), different RAM speed, and last but not least, using different laptop chassis with different cooling and firmware. Also, I have no idea if they have the same power limitations.

That's not important, all I'm saying is that one single SKU can't tell us everything. Despite the minimal differences, one is 53% faster than the other, one is 29% faster. Yes, 45W is quite a lot more than 28W, I know. Keep focused on the two 155H models instead.

Chances are that something faster than the S 14 exists, just like the Magicbook is so much faster than the Galaxybook.

View attachment 365821

My point is all about promises and bold statements vs. reality. If this result delivered AMD, everyone would spit shit on them just like they did when Zen 5 came out. Intel is not saint either and should face a reality that they delivered something else than what was expected (because of stupid megalomanian bold statements).

The fact is, Intel does this always. Just like Intel says that their 18A is healthy and already so smooth that it can run some RAM speeds out of the box and boot stuff, yet Broadcom sees things differently. Newest rumors speak about intel having set their goal too high to achieve, meaning the yield rate is very bad, like what's Samsung currently facing with its 3nm node.

Arrow lake was meant to be fully made on 20A and supposed to be already here (delivered in H1 2024), the reality is TSMC 3nm and October 2024.

Man, if I don't know what will come, I rather don't make bold statements again, and again, and again, and again.

There's a saying: Pride comes before a fall. I don't want for Intel to fall, but it would served them right.

 

[SL2]

So I take it that you missed my point by far?

Your "reality" is ONE laptop, and just like when Meteor launched, something faster might show up that we don't know about.

 

[LittleBro]

And that something new will finally beat all the expectations? So it's not about manufacturers statements vs. reality at the time of the product release? There's always a flagship-like product that comes along with a new PC hardware. Chip manufacturers partner with product makers to deliver the flagship to demonstrate new chip's capabilities. Are you implying that this time is yet to come for Lunar Lake?

 

[SL2]

And that something new will finally beat all the expectations? So it's not about manufacturers statements vs. reality at the time of the product release? There's always a flagship-like product that comes along with a new PC hardware. Chip manufacturers partner with product makers to deliver the flagship to demonstrate new chip's capabilities. Are you implying that this time is yet to come for Lunar Lake?

I'm not implying shit.

We can't conclude much from on SINGLE model like you do.

I'm not saying the review is wrong. Maybe this review is 99.9% correct on what that IGP can do. Maybe it isn't.

 

[ThomasK]

In gaming Ultra 9 285K will be best on the market.

You'll be sorry to hear that Intel itself admitted Arrow Lake is worse than Raptor Lake for gaming.

Now stop drinking your kool aid.