Intel Core Ultra 9 285K

[MaMoo]

Was looking forward to Beast Lake. That would have been the future.

 

[Ayhamb99]

AMD has kind of had a lackluster launch of it's latest series, but most everyone anticipates and expects the latest X3D parts should pick up on a positive note just where AMD generally left off X3D. It should resolve the bigger CCD and X3D latency complaints as well. Not a great time for Intel to blunder, but AMD had a small stumble in terms of underwhelming with it's recent launch to some extent.

The issues with the CCD latency was addressed with AGESA 1.2.0.2 which lowered inter-core latency by a lot, as for the cache latency i guess we're going to have to wait until the Zen 5 X3D chips to come out. Zen 5 wasn't bad in any metric it just fell victim to overhyped marketing with exaggerated performance gain claims, which led to disappointment.

It feels a good bit like Intel made multi-threading worse overall from previous generation with removal of HT and doing nothing to increase E cores in it's absence. Like overall thread count is worse between 265K and 14700K and also in more than a handful of scenario's even the 285K ends up lower than a 14700K due to fewer threads. For a new product launch I wouldn't call that great to have the previous generation i7 beating the next generations i9. To quote Steve from Gamer's Nexus in a previous Intel review "Waste of Sand" is pretty close to my feelings on this one.

Generally in the total relative average though for applications performance the 285K and 265k were ahead of the 14900K and the 14700K despite the Arrow lake skus having less threads, especially in rendering. Many of the scenarios where Arrow lake falls behind I suspect is not due to less threads but due to poor optimisation or bugs or something. The MySQL results for example are all over the place, which makes me wonder.

@W1zzard does MySQL usually have weird results like that with each new generation, The 9700X being faster than the 9950X and 7950X but the 9900X somehow being slower than the 5950X makes no sense.

Spoiler

 

[InVasMani]

They both suck, what you gain with more E cores in MT performance you lose in the ratio of cores that are capable of higher ST performance.

Intel didn't add more E cores they dropped HT tweaked cache and iGPU and slightly adjusted the frequencies and voltages then packaged it all on a new socket. I think the next generation on newer socket will Intel is liable to drop 2-4 P cores and insert more E cores while closing the game further between E core performance relative to P cores. The remaining P cores I think will probably get a bit of cache improvement. Perhaps they'll bring AVX-512 back.

I wonder if they might shift a instruction set two off the P cores onto the E cores. In fact since they also have the island E cores some could be shifted to that as well. It could further separate core types a bit into more targeted roles. I think slightly deviating x86 chips into more spread out diversified ASIC like chips is probably the right path forward. It's a matter of inserting enough of the right targeted chips in the right places to meet peoples needs. I'd say X3D is good example of technology that's pretty squarely aimed some particular use case audiences.

The way things are going I wouldn't be surprised if Apple ends up buying out Intel in a hostile take over.

The issues with the CCD latency was addressed with AGESA 1.2.0.2 which lowered inter-core latency by a lot, as for the cache latency i guess we're going to have to wait until the Zen 5 X3D chips to come out. Zen 5 wasn't bad in any metric it just fell victim to overhyped marketing with exaggerated performance gain claims, which led to disappointment.


Generally in the total relative average though for applications performance the 285K and 265k were ahead of the 14900K and the 14700K despite the Arrow lake skus having less cores, especially in rendering. Many of the scenarios where Arrow lake falls behind I suspect is not due to less threads but due to poor optimisation or bugs or something. The MySQL results for example is all over the place.

Decompression is defiantly seems to be thread impacted. That kind of coincides with the gaming situation a bit too plenty to decompress and games today are certainly more thread aware today than a decade ago. Overall games have grown more dynamic and can leverage more threads for more things I think is just a reality. They aren't all programmed well for that, but that's another matter which probably game engines will heavily address. I'm not too sure on MySQL maybe compression/decompression is also having some sort of impact and in different manner than the alternatives to it.

 

[trparky]

The way things are going I wouldn't be surprised if Apple ends up buying out Intel in a hostile take over.

I actually see Samsung buying Intel more so than Apple.

 

[InVasMani]

I actually see Samsung buying Intel more so than Apple.

Seems plausible enough as well. It just seems like if Intel's stock price plummets beyond a certain point a hostile take over becomes pretty plausible.

 

[trparky]

Seems plausible enough as well. It just seems like if Intel's stock price plummets beyond a certain point a hostile take over becomes pretty plausible.

But what would Apple do with Intel? OK, they might take them for their fabs, but it would take years and billions of dollars to get them up to the point where TSMC is today. They have no use for the x86 side of the house since Apple is focused on ARM so then what? What would they do with x86? Sell it to AMD? Ehh, I don't like that idea one bit.

 

[Ayhamb99]

The way things are going I wouldn't be surprised if Apple ends up buying out Intel in a hostile take over.

I actually see Samsung buying Intel more so than Apple.

Seems plausible enough as well. It just seems like if Intel's stock price plummets beyond a certain point a hostile take over becomes pretty plausible.

AMD was in a similar/sort of worse state after Bulldozer than Intel is right now and yet they still pulled through. Yes things looks bad and yes the management is mainly to blame for their incompetence, but regulators will not allow Intel to just get bought out like that, especially the U.S gov who would never allow a giant American corp like Intel to just collapse and get taken over.

What absolutely needs to happen if Intel wants to survive is for the management/executive board to get completely fired for their utter incompetence and arrogance. Intel's struggles in the future will probably be used as a prime example of what happens to a company when people with MBAs and zero expertise in the company's field are put in charge.

 

[Random_User]

The saddest thing is that, Intel now has more or less parity (same silicon manufacturer- TSMC), and still loses in gaming, while using superior node. Imagine this running at whatever their own "superior" node.
Of course, gaming is not everything. But the cost, solely for the production still leaves a bitter taste. The only plus is, that this time the production users can use this safe, as TSMC didn't report any manufacturing issues.

 

[csendesmark]

Not for what I do. I read these reviews and went straight to the Science and Physics tests, because that is what matters to me most. Gaming is a secondary concern as ANY CPU that performs well with these kinds of tasks will do games VERY well.

Intel Core Ultra 9 285K Review

Finally! Intel's new Arrow Lake architecture is launched. The new CPUs are full of design changes, like removal of Hyper-Threading, new Lion Cove P-Cores, an improved Thread Director and more. In our review we got surprising results that were both impressive and disappointing.

www.techpowerup.com

In these tests, the 285K kicks the crap out of most of the competition and handily beats the 9950X. The 9950X3D likely won't do much better if at all. We'll see though.

I see, fair point.
But I will ask if there is a way to utilize your GPU on those stuff is possible or not?
And if yes, how much advantage those have over the CPU-s?

 

[Garrus]

Aggregated performance is nearly meaningless, as the review focuses on various power user/workstation/professional workloads, and as I mentioned many of which are relevant for development and creative works are cases where Arrow Lake outshines the competition. And likewise, many other workloads Zen 4/5 with 12/16 cores excel by a good margin, many of which are more batch workloads. And this goes to show that for any "prosumer", aggregated performance results are largely useless, as the variance between tests aren't a few percent, but can be easily 30-50% in either direction in many cases. And there is no point in buying the perfect hardware for a use case you don't have, which is why reviews should rather say for workloads like A, B and C, this product is clearly better, and for D, E, F and G that is better, rather than an aggregated score.


Did you even read my post?
"within 2-4% in 1440p/4K with a high-end GPU"
Which is very accurate according to TPU's review.
Keep in mind that no one except a few forum users and YouTubers run a RTX 4090 at 720p on low in order to create arificial bottlenecks.

Also keep in mind that Arrow Lake is at a disadvantage here, running "underclocked" memory while Zen 4/5 is overclocked from 5200/5600 MHz respectively. It only matters a tiny bit, but then again, the aggregate results have only tiny margins.
Additionally, as said the results are skewed at bit from outliers, which is a problem with averages, so depending on the game selection the variance will be even less.

So I stand by my statement that it's pratically a draw in most cases. For most current games, any of the faster Zen 4, Zen 5, Raptor Lake or Arrow Lake CPUs will perform virtually the same, especially if paired with something like a RTX 4060 or 4070, and running a stock system (not some extreme OC to set records). In many ways this is a good position to be in, since you can pick the CPU that fits your other requirements, whether those are specific application performance, platform IO, or just whatever is the better deal at that point, and still get top class gaming performance.


A lot of sense in your post except for suggesting to put things on the back. This causes more problems than it solves, needs different cases for access and airflow, makes cases wider, etc.
Just abandoning M.2 for desktop is the better solution; provide all CPU and chipset lanes as PCIe slots, which would make motherboards cheaper at time of purchase, and give much more flexibility down the road. Buyers now need to pick motherboards very carefully, accounting for which IO they might need a few years from now, and if they pick the wrong one, it will effectively shorten the longevity of their system.


The massive metal blobs on the latest Intel and AMD platforms are absurd.
As you suggested, putting SSDs on PCIe cards makes this easily solvable, as a tiny heatsink with good long fins easily cools far more than giant metal blobs, and in extreme cases a tiny fan will do the rest.
But having these massive metal blobs (underneath the graphics card) just absorb heat for a short burst, and then very slowly dissipate it. This doesn't just shorten the lifespan of hardware, it also leads to heavy throttling.
It's about time we get some common sense motherboards with no BS and an affordable price, instead of dozens of terrible gimmicky "high-end" boards, many of which are very expensive and still doesn't fully expose the features of the chipset. This is the kind of nonsense we get when designers and marketing people try to do engineering.

oh dear you think losing by 13 percent is a draw, ok, well, you do you almost $1000 bucks Canadian to lose by 13 percent to a CPU that costs almost half as much money and uses half the electricity also

13 percent is the average, many games are 20 percent slower...

Wins both single and multi in rendering. E-cores are 50% more productive. And that's what it was created for.
Apparently not for gaming. since the memory controller and Phy were separated from the CPU tile unlike Lunar lake and that is so easy to fix. But in the current state it's DOA for esports.

View attachment 369100

It doesn't win in ST and MT. It wins at Cinebench. You just ignored all the other tests in Techpowerup's review. They outright summarize application performance at the end of the review as around 1.5 percent faster than the 14900K and slower than the 9950X by 3 percent. Uses much more power and causes random blue screens and crashes.

I'll wait for a re-review 2 months from now after the X3D parts are out. This is not a day one buy for sure.

 

[igormp]

I see, fair point.
But I will ask if there is a way to utilize your GPU on those stuff is possible or not?
And if yes, how much advantage those have over the CPU-s?

Consumer GPUs lack proper fp64 support, so they're awful at high-precision tasks.
Lots of scientific/hpc stuff requires high precision, so...

 

[lexluthermiester]

and still loses in gaming,

It breaks even in gaming and wins in nearly everything else.

Are people blind or is reading comprehension failing? Or maybe I'm missing something..

But I will ask if there is a way to utilize your GPU on those stuff is possible or not?

That's the balance I'm aiming for. The right CPU paired with the right GPU, or in my case GPGPU + GPU.

 

[trparky]

Intel's struggles in the future will probably be used as a prime example of what happens to a company when people with MBAs and zero expertise in the company's field are put in charge.

Boeing has entered the chat.

 

[Pumper]

I actually see Samsung buying Intel more so than Apple.

The chances of US government allowing Koreans to buy a strategically important American company are exactly 0.

 

[W1zzard]

But I will ask if there is a way to utilize your GPU on those stuff is possible or not?
And if yes, how much advantage those have over the CPU-s?

It depends a lot on the task you are computing. GPUs are really good at each of the thousands of cores doing the same operation on a large set of data. CPUs are the opposite and can do lots of mixed operations. Also you have to consider that many smaller problems aren't worth running on a GPU, or are very complex (developer time/developer training) to run on a GPU, while it's super easy on a CPU.

Also, GPUs have relatively little memory and adding more memory means buying an expensive pro-level GPU. CPU memory is almost free.

For Comsol specifically: https://www.comsol.de/support/knowledgebase/866

 

[Smartcom5]

GN specifically set their tests up to try to account for this.

Yes, thank God for that – Trying to detect cheating behaviour. ASUS at it again, just like when back in the days came up with M.C.E. …
It was intentionally, since ASUS' engineers (who at least had the decency to confirm it having done so) actually engineered it precisely that way,

They dit it, to re-route a good portion of the actual power-consumption (but not too much of it, to make anyone too suspicious…) going literally another route, and purposefully being not measured and left out – All to deliberately make it look that Arrow Lake would be way more efficient than what it actually is, at least on ASUS' boards.

All the while knowing full well, that the majority of reviewers and outlets only either measure actual package-power (which could be secretly accounted for for the actual missing percentage via BIOS-shenanigans; substract ~50W generally) or use clamps on the cables for EPS12V.

---

Though it's quite funny (and telling…) to see, how basically no-one really sees a need for actually making any greater fuss out of it – They all take another board, and that's it.

When AMD had their #PowerGate on their RX480 back then, it was made a sh!tstorm bar none by basically every bigger outlet. Since virtually everyone was implying, that AMD having done it fully on purpose (despite having exactly none evidence to claim such in the first place), when it also genuinely could've been a actual software-bug in BIOS.
It was fixed like in single-digit days and AMD took full responsibility for it anyway (issuing that it really shouldn't work that way).

Yet no-one really was eager to draw the full picture – There were already a number of graphics-cards which happened to do precisely the same (over-taxing it PCIe-slot's 75W by far), and it were in particular graphics-cards from of all things nVidia which took it the most far on that.

The similarities of one-sided outrage is really striking. Since if the very same would've happened on AMD-boards … My oh my, the outcry!
Disgusting hypocrisy on this again, since it's just Intel. They can do the nastiest sh!t and no-one really cares. A good part of reviewers doesn't even mention it.

 

[b1k3rdude]

This is NOT a positive -

• NPU included - Oh joy so 24H2 will try to install Co-Pillock on my PC - HARD PASS!!!!!

And these are neither here nor there to the point of 'who cares' -

• Integrated GPU - This is normal now, so why mention this AT ALL?
• Support for Thunderbolt 4 & 5 - who other than Mac users, use this..?
• 3 nanometer production process - Serously? why would the avg user give two sh*ts about this..?

And these are misleading -

• Excellent performance in heavy multithreaded apps - in Synthetic tests, for now.
• Good energy efficiency - only compared to rapter lake, Arrow lake is still crap overall.
• PCIe Gen 5 SSD without compromising GPU bandwidth - And AMD dosent..?
• Good memory support, well over DDR5-8000 - faster ram dosent make this CPU any better, so this is irellevent.

 

[RootinTootinPootin]

Good memory support, well over DDR5-8000 - faster ram dosent make this CPU any better, so this is irellevent.

it does, but if you can maintain Gear 2, else its really gonna be slower due to where the IMC is located on the Tile Layout. That is not anything a firmware or BIOS update can fix.

 

[Caring1]

But what would Apple do with Intel? OK, they might take them for their fabs, but it would take years and billions of dollars to get them up to the point where TSMC is today. They have no use for the x86 side of the house since Apple is focused on ARM so then what? What would they do with x86? Sell it to AMD? Ehh, I don't like that idea one bit.

You mean sell it BACK to AMD if that is even a thing.
Intel only licensed X86 from them.

 

[FlyingHopes]

The funny thing, even the NPU performance isn't good as its CPU and GPU side almost. The reason of increase price is this NPU, but I'm wondering are they optimize this little NPU for workloads for better use instead of using CPU and GPUs. Even NPUs of Core Ultra-H series mobile CPUs have better performance than this, if I not remember wrong.

 

[LittleBro]

Obviously 24H2, since it has been in the public channel for a while now. Testing 23H2 with Intel was a courtesy, since it is a version that is disappearing.



It makes no sense. Imagine if one platform had a 3-channel controller and another a 2-channel controller (which has happened in the past). Are you essentially saying to use 2 sticks even in the 3-channel platform to be fair? That doesn't make sense.

What does make sense is to test each platform under their ideal conditions and then point out that 8000 MT/s RAMs cost more, so you also have to consider their price in the cost/performance ratio.

You did not get my point. Both Zen 5 and Arrow Lake are 2-channel memory destups. AMD can do 8000 MT/s on Zen 5 as well as Intel. Then if Intel gets tested with 8000 MT/s, AMD should be retested as well. AMD's worse scaling per memory speed shows worse IMC than has Intel, this is perfectly okay. Anyway, testing both at same speed, be it 6000 or more MT/s, has a meaning - it disregards impact of the memory speed, which will show pure generational improvements.

 

[trparky]

The chances of US government allowing Koreans to buy a strategically important American company are exactly 0.

Yeah, but do you see Apple doing it? Not a chance! That's why I'm saying that there's more of a chance that Samsung would buy them than Apple. I'm not saying that the US government would allow it, I'm just saying that Apple probably wouldn't want a thing to do with Intel.

You mean sell it BACK to AMD if that is even a thing.
Intel only licensed X86 from them.

Uh... the only thing Intel licensed from AMD was x86_64.

 

[oxrufiioxo]

Yeah, but do you see Apple doing it? Not a chance! That's why I'm saying that there's more of a chance that Samsung would buy them than Apple. I'm not saying that the US government would allow it, I'm just saying that Apple probably wouldn't want a thing to do with Intel.

Uh... the only thing Intel licensed from AMD was x86_64.

The only reason I could see apple acquiring them is for their foundries and the uncertainty with TSMC being in Taiwan.

Now is that enough to make Apple want to purchase Intel probably not but there are only a hand full of companies that could afford to aquire them it could also have more to do with stopping a merger of Qualcomm and Intel which competes with both Samsung/Apple.

I agree though I can't imagine Intel is all that desirable right now other than their foundries which while not as good as tsmc are still much better than any other alternative.

It's crazy that AMDs market cap is 2.5x intels right now.

 

[AnonymousGuy767]

Intel needed to release this 3 years ago. Now I think there's not enough runway left to allow a "first generation" type reset to mature, which the 285K is. Plus the turmoil internally isn't good either. A lot of reviews indicating that Windows isn't ready for it at launch...that's stuff that having tenured employees with good understanding of "small" details would take care of.

 

[Oberon]

It breaks even in gaming and wins in nearly everything else.

You misspelled "loses outright in gaming and barely has parity in productivity."