News Core Ultra 9 285K is slower than Core i9-14900K in gaming, according to leaked Intel slide — Arrow Lake consumes less power, though

[TheHerald]

T

That's the point. 9950X is the top end Zen 5 CPU, i9-14900K is the top end Raptor Lake. It's a fair comparison.


Not if we're talking about gaming, which is the subject of this article:

​

​

Source: https://www.tomshardware.com/pc-com...built-in-benchmarks-versus-custom-game-scenes​

​

The 9900X price will come down, but the i9-14900K is probably about as cheap as it'll get. For someone buying a CPU today, the latter represents a better value. If we're looking down the road, then i9-14900K will have more to fear from the 9950X.

Was true in the past but lately intel seems willing to drop prices further. 12900k for 265$ for example.

 

[bit_user]

Was true in the past but lately intel seems willing to drop prices further. 12900k for 265$ for example.

It's a smaller die than Raptor Lake, so who knows?

 

[YSCCC]

You said ryzen 9000 is already faster than the 14900k, obviously you weren't talking about gaming. In which case it's irrelevant, most of the 14th gen product stack is much much faster than the zen 5 product stack, and even if ARL gets a performance decrease it will still remain much faster than ryzen 9000.

The 7800x 3d is irrelevant, it's never stealing sales from i9s or ryzen 9s. Nobody is between an 8 core megaslow cheap and a 9950x. Come on now, the performance discrepancy between these chips is huge. If you are looking for anything like a 9950x, 7950x, 7950x 3d, 14900k etc. the 7800x 3d might as well be an i3. Doesn't matter. It might steal sales from an i5, sure, but i9s / r9s etc are way outside it's league to be even a contender.

Normal ppl will read as top 9000 sku is faster than the 14900k, gaming or not, and not implying whole zen5 is faster than 14900k, not specifying the 9950x is because this topic is on ARL, no point to detail here, if you misunderstood, this is the clarification that I mean the top vs top

And of course 7800X3D is relevant, in pure gaming it is still the fastest, so if gaming is priority, the 9000 series coming X3D, especially if the 9900X3D or 9950X3D is as rumoured being will full 3D v cache and not only on one CCD, it will be very relevant in cache dependant workloads, and even with current 7800X3D, it is preferred to 14900k if one is only into gaming and not rendering. Which a lot of ppl is already building it for.

 

[TheHerald]

Normal ppl will read as top 9000 sku is faster than the 14900k, gaming or not, and not implying whole zen5 is faster than 14900k, not specifying the 9950x is because this topic is on ARL, no point to detail here, if you misunderstood, this is the clarification that I mean the top vs top

And of course 7800X3D is relevant, in pure gaming it is still the fastest, so if gaming is priority, the 9000 series coming X3D, especially if the 9900X3D or 9950X3D is as rumoured being will full 3D v cache and not only on one CCD, it will be very relevant in cache dependant workloads, and even with current 7800X3D, it is preferred to 14900k if one is only into gaming and not rendering. Which a lot of ppl is already building it for.

Gaming is a priority for me but I'm not touching an 8core chip. Gaining 1% if even that at normal gaming and losing half the mt performance vs a 9950x is a ridiculously silly trade to make.

You know this yourself, that's why you have a 14900k. With the 900k money you could have bought a 7800x 3d (that according to you is faster) and a new motherboard on top of that, yet you didn't. Your arguments might disagree with me but your buying decisions do not.

 

[TheHerald]

It's a smaller die than Raptor Lake, so who knows?

Yeap, was thinking about that as well, but realistically are they selling the 12900k at cost right now? Most likely not. Time will tell.

 

[YSCCC]

You know this yourself, that's why you have a 14900k. With the 900k money you could have bought a 7800x 3d (that according to you is faster) and a new motherboard on top of that, yet you didn't. Your arguments might disagree with me but your buying decisions do not.

Nope, you put your word in my mind again, I bought the 14900k is simply, I got the whole system done, I want an good enough upgrade to last for 3 more years, AND no need to re-do the whole system and no need to get another copy of windows, the windows alone is $200 and what I got back then was a OEM ver. where changing board will need a new license. That was why the 14900k is close enough and saved a ton of money and building time including reinstalling every single software trouble if a system swarp is to be done. If a whole rebuild was to be done, I wouldn't upgrade altogether until like 2 gen later and the 12700KF is really unbearably slow. Plus I said a dozen of times, I used to believe intel have better longevity and not the whole mess of degradation, so the slower performance and higher power consumption was ok.

If you would buy me a brand new 7800X3D with motherboard and a genuine Windows (OEM or box), I would just give my 14900K to you, would you?

Edit: Just in case you forgot, I bought the 14900k on release date, where the degradation issues was not known and thought after 13th gen, it should be rock solid and safe on unlimited power so I can OC in future if I wanted, plus no need to replace XMP ram to EXPO ram set. If 14900k was stable and issue free as intel used to be, it was indeed very competivtive at the point of time as MT performance was slightly better than 7950X was at the time, and gaming isn't too far behind the X3D as to make it completely obsolete, now the whole issue have unravelled and it becomes a really bad gen.

Back onto topic, now the ARL was actually having gaming performance regress and gaming, and rendering performance improvement as you've said for Zen5 vs Zen 4 of 1x% being "pathetic", (e.g. Puget bench they have basically zero improvement), it is not a good launch/bright future for Intel.

 

[TheHerald]

Edit: Just in case you forgot, I bought the 14900k on release date, where the degradation issues was not known and thought after 13th gen, it should be rock solid and safe on unlimited power so I can OC in future if I wanted, plus no need to replace XMP ram to EXPO ram set. If 14900k was stable and issue free as intel used to be, it was indeed very competivtive at the point of time as MT performance was slightly better than 7950X was at the time, and gaming isn't too far behind the X3D as to make it completely obsolete, now the whole issue have unravelled and it becomes a really bad gen.

Back onto topic, now the ARL was actually having gaming performance regress and gaming, and rendering performance improvement as you've said for Zen5 vs Zen 4 of 1x% being "pathetic", (e.g. Puget bench they have basically zero improvement), it is not a good launch/bright future for Intel.

That is kinda exactly what im saying. I underlined it for emphasis. It's so much faster than the 7800x 3d in anything but gaming that the 7800x 3d wasn't even on your radar. Which is what im saying, people looking at the 9950x, 7950x, 14900k or the new i9 don't particularly care about the 7800x 3d or the 9800x 3d. It's a console replacement CPU, it's not an enthusiast tier chip.

EG1. I would absolutely trade a 7800x 3d for a 14900k, lol.

 

[YSCCC]

That is kinda exactly what im saying. I underlined it for emphasis. It's so much faster than the 7800x 3d in anything but gaming that the 7800x 3d wasn't even on your radar. Which is what im saying, people looking at the 9950x, 7950x, 14900k or the new i9 don't particularly care about the 7800x 3d or the 9800x 3d. It's a console replacement CPU, it's not an enthusiast tier chip.

EG1. I would absolutely trade a 7800x 3d for a 14900k, lol.

You don't read again and that what I said is NOT what you said.

What I said was simple, the motivation of MY own upgrade decision was based on the upgrade will be desirable only if I don't need to rebuild everything and get a new copy of windows, which could easily double the price of the upgrade for expected 3 extra years of usable service life. IF intel have had a 7800X3D type of processor, I would've gone that way instead, but no, intel only have i5,i7 and i9 with no X3D, so based on what I could do in a 3min swarp of a CPU and gain extra 1Ghz and 12 more e cores with more cache, plus no need for the permanently deformed IHS to have one core throttle during spikes, it was worth it to spend $600 (back in the days in local currency) to upgrade in a blip and not half a day rebuilding everything and create a ton of e-waste.

If the 14900k cannot be drop and run directly, I wouldn't have even consider it and wait 2 more gen. That is what drop in upgrade is temptative to quite a lot of users.

Now that was without the huge mess of degradation where they supposedly have finally fixed by microcode limitation of voltages. Assume that intel didn't have this huge disaster for RPL to boot, they are competative to AMD, nobody said 14th gen is not competative in pure performance metric, the RPL is bad purely because it failed miserably on reliability front.

And back to topic and not derail chief, for new Ultra 9 285K, if this leak is true (which could always be not true, time will tell), it is SLOWER than 14900k in gaming, while the competition Zen 5 9950X is faster than Zen 4 7950X, which is kind of on par or quicker than 14900k depends on what you throw at it, 9950X is likely faster in gaming than upcoming 285K, and X3D is likely even faster than it in gaming, and hack the figures are from Intel itself, which both camp traditionally showed figures higher than real reviews, the picture don't look to be really competative to Intel.

P.S. And for exchange, it is not changing a 7800X3D with a 14900K, is changing 7800X3D+board+expo ram and a new windows with a 14900k+XMP ram, which is what I have to choose when I decided to in slot upgrade, which you accused that I personally won't choose a 7800X3D over 14900k, would you do so?

 

[TheHerald]

P.S. And for exchange, it is not changing a 7800X3D with a 14900K, is changing 7800X3D+board+expo ram and a new windows with a 14900k+XMP ram, which is what I have to choose when I decided to in slot upgrade, which you accused that I personally won't choose a 7800X3D over 14900k, would you do so?

Yes, you don't need a new windows and even if you do the costs is 20$, a 7800x 3d + a mobo for it is cheaper than the 14900k itself, so it's basically 0 cost. Actually I'd even make money back by selling the old motherboard. So yes, if I thought the 7800x 3d is better than the 14900k, i'd do it. I don't think it is, so I wouldn't.

You can resuse your windows license if you register it to your account, and frankly you can fine a w11 pro license for 7,99$. A legit one. So yeah...if 8$ is what stopped you from "upgrading" to the 7800x 3d instead of the 14900k, just send me a pm next time. Ill fork out for you.

And back to topic and not derail chief, for new Ultra 9 285K, if this leak is true (which could always be not true, time will tell), it is SLOWER than 14900k in gaming, while the competition Zen 5 9950X is faster than Zen 4 7950X, which is kind of on par or quicker than 14900k depends on what you throw at it, 9950X is likely faster in gaming than upcoming 285K, and X3D is likely even faster than it in gaming, and hack the figures are from Intel itself, which both camp traditionally showed figures higher than real reviews, the picture don't look to be really competative to Intel.

This is all over the place. You are saying that the 9950x is faster than the 7950x in gaming, and that the 7950x is on par or quicker than the 14900k? That's just absolutely unequivocally wrong. I just check techpowerup, computerbase de and hwunboxec All of them disagree with you, the 14900k is on par or faster (for example in hwunboxed) with the 7950x 3d. The normal 7950x and the 9950x are both slower than the 14900k. So arrow lake will be faster than both the 7950x and the 9950x in games, no questions asked.

 

[YSCCC]

Yes, you don't need a new windows and even if you do the costs is 20$, a 7800x 3d + a mobo for it is cheaper than the 14900k itself, so it's basically 0 cost. Actually I'd even make money back by selling the old motherboard. So yes, if I thought the 7800x 3d is better than the 14900k, i'd do it. I don't think it is, so I wouldn't.

You can resuse your windows license if you register it to your account, and frankly you can fine a w11 pro license for 7,99$. A legit one. So yeah...if 8$ is what stopped you from "upgrading" to the 7800x 3d instead of the 14900k, just send me a pm next time. Ill fork out for you.

Nope, I said I used an OEM win11, which is invalid to major hardware upgrade, which is, according to MS, changing board, and in socket CPU replacement is fine.

Noe it is not $20, a genuine windows 11 license is $200, not $20, those CD-key.com etc. arn't official dealer of windows, and I don't use those way off MSRP stuffs from suspicious sources, not because that it won't work, but I support the software company creating it, that is called ethics.

This is all over the place. You are saying that the 9950x is faster than the 7950x in gaming, and that the 7950x is on par or quicker than the 14900k? That's just absolutely unequivocally wrong. I just check techpowerup, computerbase de and hwunboxec All of them disagree with you, the 14900k is on par or faster (for example in hwunboxed) with the 7950x 3d. The normal 7950x and the 9950x are both slower than the 14900k. So arrow lake will be faster than both the 7950x and the 9950x in games, no questions asked.

You are checking the results from the power unlimited RPL, which, we all know, fries itself to oblivion.
Check individual titles like shown in the GN suite, they are trading blows in some, and as always some favours(optimize) better than the other.

https://gamersnexus.net/cpus/intels-300w-core-i9-14900k-cpu-review-benchmarks-gaming-power

And say in HW unbox, the 7800X3D and 7950X3D is always faster than 14900k, which is supposedly faster than the Ultra 9.

 

[TheHerald]

You are checking the results from the power unlimited RPL, which, we all know, fries itself to oblivion.
Check individual titles like shown in the GN suite, they are trading blows in some, and as always some favours(optimize) better than the other.

https://gamersnexus.net/cpus/intels-300w-core-i9-14900k-cpu-review-benchmarks-gaming-power

And say in HW unbox, the 7800X3D and 7950X3D is always faster than 14900k, which is supposedly faster than the Ultra 9.

No im not. Power limited or unlimited does no difference in gaming.


And no, in HWunbox the 7800x 3d and the 7950x 3d isn't always faster than the 14900k. This is their latest review, the 14900k is faster than the 7950x, the 7950x 3d and the 9950x. You are the first person claiming that the 7950x is on par or faster with the 14900k in gaming. That's insanely wrong. The 7950x is on par with 12th gen.

View: https://www.youtube.com/watch?v=43DFYvOoRhY

 

[ManDaddio]

I think if the CPU performs no worse than the best of 14th gen, that is good. It's a new design and efficiency. That's a good move for Intel.

I was hoping not to have to move to Ryzen. Maybe there is hope down the road.

I had to buy a 7900x because I had issues with my 13700k. After getting a new chip (13700k) and adding the mitigations I'm still more happy with my 13700k versus the 7900x.

I'm still going with Intel if they can snap back.

 

[ManDaddio]

Only matters to the rich who upgrade every cycle for no reason. I, one of the vast majority, am upgrading from an older gen, 11th gen to be precise so it doesn't matter will still be a huge upgrade. I'll just be waiting for the Ultra 9 vs X3D comparisons when both are released.

Traditionally there was a reason. But things are slowing down progression wise. At least right now.

I'm glad I upgraded every cycle.
970 to 1070/1080 to 2080ti to 3090 to 4090.

They were all relevant buys. The FPS and graphical gains were legit.

 

[YSCCC]

I think if the CPU performs no worse than the best of 14th gen, that is good. It's a new design and efficiency. That's a good move for Intel.

I was hoping not to have to move to Ryzen. Maybe there is hope down the road.

I had to buy a 7900x because I had issues with my 13700k. After getting a new chip (13700k) and adding the mitigations I'm still more happy with my 13700k versus the 7900x.

I'm still going with Intel if they can snap back.

Agree that It's a new design and efficiency. That's a good move for Intel., which, IMO, should be the 14th gen and would be a big success if the RPL refresh didn’t happen and this is the release.

For the snap back part, I believe Intel need to snap back in reliability/reputation part more than performance or efficiency part, which is much more difficult than simply outperforming competition as you can use benchmarks to prove you aren’t lacking behind in performance or efficiency, but reliability takes time, and that isn’t a short time

 

[thestryker]

Not if we're talking about gaming, which is the subject of this article:

​

​

Source: https://www.tomshardware.com/pc-com...built-in-benchmarks-versus-custom-game-scenes​

I wouldn't use that graph as a particularly fair example since it's AMD CPUs running Expo settings vs RPL at 5600 JEDEC (I am aware it's the only reliable compiled data available likely until ARL reviews).

 

[thestryker]

That's the 14900ks, the K is at 160w average. From my experience at completely stock with no power limits lowest ive seen in games was ~90w, highest i've seen was ~200w, that's on 1080p with a 4090 and completely cpu bound. I don't think you can make it draw more than 200w, in which case the supposed 170w reduction seems like they messed up something in t heir graph. There is no way ARL is dropping to 30w, lol.

That outlier is Space Marine 2 which upon launch (don't have the game so don't know if all the patches have helped) had some of the highest CPU usage any game has seen. While I agree it seems unlikely to drop that far down it is possible the game was pushing the CPU well past 200W.

 

[thestryker]

I can't say I'm surprised by the results, but I'm somewhat wondering where the gaming bottleneck happens to be. AMD has had some regressions in gaming performance with Zen 5 compared to Zen 4 despite Zen 5 clearly being the superior technology. It seems like some of that might be going on with Intel as well since the IPC is clearly higher with ARL, but still has outliers on gaming performance.

If Intel has knocked down the high power consumption outliers for gaming/light usage that alone would be a huge improvement. I've not minded minimum or maximum power consumption with Intel 13-14th as they're easy to manage it's the in between that is a lot harder.

Hopefully Tom's will release the entire ARL slide deck when the announcement embargos are up tomorrow like AnandTech always did. Will be interesting to see what Intel's expectations are and then in a couple more weeks how they've actually panned out.

 

[YSCCC]

I can't say I'm surprised by the results, but I'm somewhat wondering where the gaming bottleneck happens to be. AMD has had some regressions in gaming performance with Zen 5 compared to Zen 4 despite Zen 5 clearly being the superior technology. It seems like some of that might be going on with Intel as well since the IPC is clearly higher with ARL, but still has outliers on gaming performance.

If Intel has knocked down the high power consumption outliers for gaming/light usage that alone would be a huge improvement. I've not minded minimum or maximum power consumption with Intel 13-14th as they're easy to manage it's the in between that is a lot harder.

Hopefully Tom's will release the entire ARL slide deck when the announcement embargos are up tomorrow like AnandTech always did. Will be interesting to see what Intel's expectations are and then in a couple more weeks how they've actually panned out.

For Zen 4 and 5 it seems gaming is kind of limited by core parking bugs/issues, where in some games the like of 9700X will be faster than 9900X and 9950X, kind of make sense in some way if the threads need to exchange info to complie a frame, then the communication protocol between chiplets will be a bottleneck, maybe that's kind of true for ARL also.

For power consumption personally I agree that although there is always an efficiency war and the lower power it drinks, better, but for years the overall power draw just skyrocketed and as GPU dominates the draw, CPU becomes less relevant, especially in high end, given it didn't fry itself up.

For the slides tought I wouldn't look into more before actual benchmark results, IIRC almost all launches showed the official promotional material are overblown, sometimes massively by both camps (plus Nvidia if GPU is thrown in also), that is what marketing BS is always doing, benchmarks, especially from the "evil hater youtubers" are of more relevance as it shows basically what could happen in worse case scenario, given they well published their system and settings, that way one can estimate what is relevant to them. (I recall some reviews where A is (much) faster than B in avg FPS, but having over 400FPS avg means nothing, while 1% lows dropped to ridiculous laggy territory, that way in real gaming the avg matters nothing, but 1% low will show as stutter).

And beside performance figures, I am definitely more longing for the boring Buildzoid probe videos to see if any crazy things will be caught, this time round I am pretty sure everyone of those reviewers will hold a microscope to try spot any issues on the Intel side, hope they really make things right at launch

 

[thestryker]

For Zen 4 and 5 it seems gaming is kind of limited by core parking bugs/issues, where in some games the like of 9700X will be faster than 9900X and 9950X, kind of make sense in some way if the threads need to exchange info to complie a frame, then the communication protocol between chiplets will be a bottleneck, maybe that's kind of true for ARL also.

I'm referring to the outliers like Starfield/CS2 (there's a few others) where the 9700X is flat out slower than the 7700/X despite having the same clocks:
https://www.techpowerup.com/review/amd-ryzen-7-9700x/18.html

 

[YSCCC]

I'm referring to the outliers like Starfield/CS2 (there's a few others) where the 9700X is flat out slower than the 7700/X despite having the same clocks:
https://www.techpowerup.com/review/amd-ryzen-7-9700x/18.html

Right, I was generally looking for quite some games the single CCD parts quite often out performs the multi CCD ones, didn't catch this though

 

[levia.dragon]

I know, I know, I may be nitpicking, but how on earth is Cinebench 2024 a content creation application, unless you mean a tool used by benchmarkers to make... content?

Cinema4D is a 3D content creation application.
Cinebench is using its rendering engine as a way of testing hardware performance in it.

 

[bit_user]

I think if the CPU performs no worse than the best of 14th gen, that is good. It's a new design and efficiency. That's a good move for Intel.

That's a fine personal opinion and value judgement, but it's hardly a recipe for commercial success.

This is 2 process nodes better than their prior desktop flagship, and if it still can't manage net gains, that's certainly troubling.

I'm glad I upgraded every cycle.
970 to 1070/1080 to 2080ti to 3090 to 4090.

They were all relevant buys. The FPS and graphical gains were legit.

There's no comparison between GPUs and CPUs, because GPUs can scale performance almost linearly with the number of cores. Smaller process nodes let them increase on two dimensions, simultaneously: more cores and higher frequencies. Also, more IPC or features like RT or AI acceleration. Furthermore, GPUs typically have a 2-year product cycle, giving more time to accumulate meaningful improvements.

As we all know, the applicability of adding more than 16 threads to a CPU is limited, especially in gaming. Scaling frequencies on CPUs also burns lots of power and IPC gains are hard. That's why CPUs typically have rather modest gen-on-gen improvements.

 

[bit_user]

Agree that It's a new design and efficiency. That's a good move for Intel.,

I have a hunch that the main way they improved efficiency was just by failing to scale frequencies as much as they'd hoped, due to a thermal bottleneck from that small CPU tile, not unlike what AMD has been dealing with since Zen 4 and especially in Zen 5.

People who use direct-die waterblocks on Arrow Lake will probably unleash some meaningful gains that put it past the reach of the 9800X3D, but now we're talking about like 0.01% of the market, if that.

 

[bit_user]

I wouldn't use that graph as a particularly fair example since it's AMD CPUs running Expo settings vs RPL at 5600 JEDEC (I am aware it's the only reliable compiled data available likely until ARL reviews).

Thanks for pointing that out. As you say, I used it because it's simply the most up-to-date data we have.

I would also point out that it's testing Raptor Lake before the final BIOS update was announced by Intel. I'm not saying the two cancel out, but there are advantages on each side of the ledger vs. what we would expect to see in newer benchmarks that don't use EXPO.

 

[bit_user]

I'm referring to the outliers like Starfield/CS2 (there's a few others) where the 9700X is flat out slower than the 7700/X despite having the same clocks:
https://www.techpowerup.com/review/amd-ryzen-7-9700x/18.html

That was data from launch day. Toms' launch day review ended up being much more consistent with current performance, due to their testing methodology working around most of the AMD/Windows bugs.

Here's Starfield data from the updated Toms review I linked above:

​

It still shows the inversion between the 9700X and 7700X, when both are tested with the same updates. However, I'll bet the 9700X was still running at 65W. They don't mention PBO and this was before AMD officially blessed running the9700X at 105W.