AMD Ryzen 9 9950X3D vs Intel Core Ultra 9 285K Faceoff

It's kind of a no-brainer isn't it? I mean, they are practically equal when it comes to productivity performance, and AMD is clearly superior for gaming, so....

I wish I had an AM5 platform, but productivity was my main concern when building my current system. And at the time, AMD was lagging behind a little bit in the productivity arena.
 
Since the 9800X3D offers similar gaming performance to the 9950X3D and the results were heavily waited toward gaming, if you had replaced the 285K in this article with the 9800X3D, I think the 9800X3D would look pretty good.

If you eliminate gaming, the 285K is $110 cheaper today on B&H and offers similar productivity performance at similar power consumption so it'd seem like the better value. Also regarding the statement, "we typically use our systems more than letting them idle for prolonged periods." I don't think that's true. Maybe if you only use your computer to game you turn it on, game, and turn it off. (If that's you, get the 9800X3D instead of the 9950X3D.) But when getting work done the CPU is rapidly going between idle and busy as you interact with it and pause to think. And a lot of people do leave the computer running while they step away from it.
 
^^ Hmm.. all valid points. For me, I work from home and my PC is in constant use, and between work and casual use, probably only sitting idle about 2 hours out of 12. Realistically, most people would be happy with either CPU, but the longevity of the AM platform appeals to me. I'm the type of person that builds one machine every 10-15 years and keeps upgrading it to keep up.
 
Also regarding the statement, "we typically use our systems more than letting them idle for prolonged periods." I don't think that's true. Maybe if you only use your computer to game you turn it on, game, and turn it off. (If that's you, get the 9800X3D instead of the 9950X3D.) But when getting work done the CPU is rapidly going between idle and busy as you interact with it and pause to think. And a lot of people do leave the computer running while they step away from it.
Well as a forum I'm sure they are 24/7/365 doing 3d rendering and once in a while 10bit video transcoding for the occasional video they are posting....
That's the only things they tested under productivity so that's what we have to believe.
 
  • Like
Reactions: rluker5
Doesn't all light use also fall under "idle" for Intel?
My PCs are usually light use and rarely run P95.
Yes, this is accurate. Also, there is an 'active idle' use case that we model and have measurements of that type of workload. Sorry for the confusion on the wording in the article. I have corrected that.
 
Last edited:
For me, when I drop that kind of cash on a CPU, it better be freaking good at gaming or i'm not interested. The Intel series of chips are trash for gaming when compared to the X3D variants. I wouldn't consider anything Intel these days. Intel tends to require new mobos with newer releases, AM5 doesn't. I can plug replace my 9800X3D with the 9950X3D.

Plus Intel did themselves no favors with their last two generations with "crashgate".
 
  • Like
Reactions: phenomiix6
Since the 9800X3D offers similar gaming performance to the 9950X3D and the results were heavily waited toward gaming, if you had replaced the 285K in this article with the 9800X3D, I think the 9800X3D would look pretty good.

If you eliminate gaming, the 285K is $110 cheaper today on B&H and offers similar productivity performance at similar power consumption so it'd seem like the better value. Also regarding the statement, "we typically use our systems more than letting them idle for prolonged periods." I don't think that's true. Maybe if you only use your computer to game you turn it on, game, and turn it off. (If that's you, get the 9800X3D instead of the 9950X3D.) But when getting work done the CPU is rapidly going between idle and busy as you interact with it and pause to think. And a lot of people do leave the computer running while they step away from it.
X3D is a gaming processor. If you're going to compare non-gaming, then you have to consider the 9950x which is cheaper than the 285K. The 9950x is also better at single core than the x3D variant shown by the charts in this article.
 
I do remember intel releasing a 6 core CPU back in 2010, so what time are you even talking about?
The Core i7 980X. At the highly affordable price of only $999, too. Aimed squarely at the mass-market it most definitely wasn't.

By comparison, the 285K and 9950X3D are downright affordable.

Then again, this was still the era where the absolute top-end of consumer CPUs was an obnoxiously terrible value proposition. Things like the Pentium 4 EE and Athlon 64 FX 51-53-55-57 etc. We all wanted them, but very few of us could afford them.
 
I've always been an Intel CPU guy...until I saw the 285k...what a disappointment. I almost had to do a full upgrade this year, but luckily I got my PC working again. Good ole Haswell-E 5960x CPU and my ASUS Rampage V Extreme Mobo, still kickin'. They're gonna have to pry my cold dead hands off of it.
 
  • Like
Reactions: phenomiix6
For me, when I drop that kind of cash on a CPU, it better be freaking good at gaming or i'm not interested. The Intel series of chips are trash for gaming when compared to the X3D variants. I wouldn't consider anything Intel these days. Intel tends to require new mobos with newer releases, AM5 doesn't. I can plug replace my 9800X3D with the 9950X3D.

Plus Intel did themselves no favors with their last two generations with "crashgate".
I take it you also think all non x3d variants are waste then as well, just to be fair?

Also I upgraded a generation going from a 12700k to a 13900kf and by the time I see a noticeable difference from the 12700k in my 4k gaming It will be longer than AM4 lasted.
The upgrade was apparently just for the entertainment of tuning a new chip. Kind of a waste of money.

I don't mind if Intel replaced their socket with every new generation. Especially if the new socket comes with improvements.
 
In the future comparisons, I think it's also fair to include the overall platform capabilities as a separate item, since "longevity" while nice to have, it may stay behind the curve for longer in features that require stronger I/O and or better memory support, etc. Whether I like it or not, Thunderbolt is still better than USB4/C in terms of capabilties and it shows. I know that it's Intel-specific and they just won't hand it over, but even with USB4 being foced in X870 platforms, it's just not the exact same and the way it's kind of bolted to the platform is, well, lame. AMD has some catching up to do there.

Also, one potential correction (maybe?): AMD was kind of fuzzy with CUDIMM support just saying "it may work", implying they haven't gone through the effort of validating them for AM5, so that could be an interesting piece for Toms to check? Maybe like some AMD peeps said "it may just work"? In any case, thanks for sticking to JEDEC speeds and timings!

Regards.
 
I take it you also think all non x3d variants are waste then as well, just to be fair?

Also I upgraded a generation going from a 12700k to a 13900kf and by the time I see a noticeable difference from the 12700k in my 4k gaming It will be longer than AM4 lasted.
The upgrade was apparently just for the entertainment of tuning a new chip. Kind of a waste of money.

I don't mind if Intel replaced their socket with every new generation. Especially if the new socket comes with improvements.

I got my first AM4 in 2017, a Ryzen 1600. That's 6-7 years depending on how you look at it. There'll be a noticeable difference before that. Especially if you plan on playing Squadron 42 which is based on a heavily modified Cryengine 3. It's very CPU bottlenecked.
 
  • Like
Reactions: Marlin1975
>Most PCs rarely truly idle; users might leave various applications open on the desktop. There's also an 'active idle' use case wherein the user does a low-load activity, such as browsing the web or watching a YouTube video. To model this behavior, we created an active idle test (second slide) with two browser windows open (one with two tabs idling on a website and another window with a 4K YouTube video stream playing). We measure this level of activity across a 15-minute timespan.

>The Core Ultra 9 285K has 30% lower idle power consumption than the Ryzen 9 9950X3D. Furthermore, the Core Ultra 9 285 K's active idle power consumption during YouTube playback is particularly notable, as it consumes 39% less power.

Finally someone started looking at more common usage scenarios. You should also be looking at other traditionally "low intensity" scenarios. Just spinning my mouse around fast will cause my 9950X3D to suck up 60-70W of power, whereas doing the same thing on Intel's Core Ultra 9 285K causes almost no difference in power consumption.
 
Same here.

Instead, i'm stuck with my Z790 motherboard for the foreseeable future.

Fortunately, 13900K is still a potent processor and i only game at 4K, so i guess i'm fine for the time being.
Actually every one is fine unless your game is below 120fps and you dont need to beat the fps for gameplay if you have 120fps. I dont understand how people look at gaming performance nowadays . you need better GPU for gaming , dont waste much on CPU , pick AMD or intel CPU , both will be fine.
 
>Most PCs rarely truly idle; users might leave various applications open on the desktop. There's also an 'active idle' use case wherein the user does a low-load activity, such as browsing the web or watching a YouTube video. To model this behavior, we created an active idle test (second slide) with two browser windows open (one with two tabs idling on a website and another window with a 4K YouTube video stream playing). We measure this level of activity across a 15-minute timespan.

>The Core Ultra 9 285K has 30% lower idle power consumption than the Ryzen 9 9950X3D. Furthermore, the Core Ultra 9 285 K's active idle power consumption during YouTube playback is particularly notable, as it consumes 39% less power.

Finally someone started looking at more common usage scenarios. You should also be looking at other traditionally "low intensity" scenarios. Just spinning my mouse around fast will cause my 9950X3D to suck up 60-70W of power, whereas doing the same thing on Intel's Core Ultra 9 285K causes almost no difference in power consumption.
This is while in a call in Discord, as I got curious by your post. Also I have no idea how ~50W is for this. When gaming the 9950X3D I see using ~80W at most. Then again, I have it with -30 curve.

View: https://imgur.com/a/SUEEhGa


I hope that worked to share the image, lol.

Regards.
 
This is while in a call in Discord, as I got curious by your post. Also I have no idea how ~50W is for this. When gaming the 9950X3D I see using ~80W at most. Then again, I have it with -30 curve.

View: https://imgur.com/a/SUEEhGa


I hope that worked to share the image, lol.

Regards.
Cores are using 7W soc is using 23w ,misc is another 10W
soc has to manage everything and especially if you run expo xmp pbo it runs beyond what it was made for and will use more power.