9800x 3d vs 12900k - Battle of the Century

[RootinTootinPootin]

I will join in the test as soon as the Tachyon arrives..will see it 1st hand if there really is voodoo magic with X3D and 4K gaming.

 

[phanbuey]

I will join in the test as soon as the Tachyon arrives..will see it 1st hand if there really is voodoo magic with X3D and 4K gaming.

Yes! let's see it.

 

[JustBenching]

I will join in the test as soon as the Tachyon arrives..will see it 1st hand if there really is voodoo magic with X3D and 4K gaming.

4k gaming? There will be 0 difference if you have anything that wasn't released a decade ago.

 

[wolf]

4k gaming?

Upscaling is something that affects that, as you draw from lower internal resolution rendering first. 9800X3D has felt like a large uplift over my 5800X3D at 4k output resolution.

I'm also yet to see a single game stuttering on the 9800X3D, granted I don't have a tuned 12900K system to try my best to do like for like testing against, I can't help but feel something is amiss. Every game so far is smooth as butter, with the caveat my chip is also tuned.

 

[phanbuey]

Upscaling is something that affects that, as you draw from lower internal resolution rendering first. 9800X3D has felt like a large uplift over my 5800X3D at 4k output resolution.

I'm also yet to see a single game stuttering on the 9800X3D, granted I don't have a tuned 12900K system to try my best to do like for like testing against, I can't help but feel something is amiss. Every game so far is smooth as butter, with the caveat my chip is also tuned.

4k used to mean "fully gpu bottlenecked" but you can do that at 1080P these days with path tracing.

Reviews should target a certain FPS - like say 120FPS with X gpu, If your goal is to average 120FPS with X gpu, will you see a % difference across X number of games. As the FPS number goes up, say you have a 240fps monitor and you want to target as close to 200 FPS as you can, then you will see Y difference.

I agree -- At 4k no one is sitting around rendering blurry native TAA at 4k if they have DLSS quality + sharpening + 30% more fps available.

 

[TSiAhmat]

not sure if this could change anything:

maybe only one of you has DLDSR enabled?

It's kinda hidden in the setting and applies in all games, if enabled over global settings.

 

[JustBenching]

Upscaling is something that affects that, as you draw from lower internal resolution rendering first. 9800X3D has felt like a large uplift over my 5800X3D at 4k output resolution.

I'm also yet to see a single game stuttering on the 9800X3D, granted I don't have a tuned 12900K system to try my best to do like for like testing against, I can't help but feel something is amiss. Every game so far is smooth as butter, with the caveat my chip is also tuned.

What kind of games do you actually see a difference from 5800x 3d ---> 9800 x3d? Cause I really can't see a difference between 12900k ---> 9800x 3d. I mean sure, with the framegraph open overwatch / cs / dota 2 is faster on the x3d but the 900k was already at or above 500 fps on those, even my monitor set to 480hz couldn't register an actual difference.

To conduct the comparisons I have I had to drop to 1080p with DLSS Ultra performance to differentiate between the chips. Haven't posted a video with all the data compiled yet cause im too busy playing POE2, but when the difference is ~15% at 1080p dlss ultra performance, oh well.

 

[wolf]

What kind of games do you actually see a difference from 5800x 3d ---> 9800 x3d? Cause I really can't see a difference between 12900k ---> 9800x 3d.

Your heavily tuned 12900K is surely a better example than stock which reviews would show, and I'd wager it's on (tuned?) DDR5 not DDR4 like my set and forget XMP DDR4 on the 5800X3D.

I can't really comment with certainty around 9800X3D vs heavily tuned 12900K, my point before was that I'm not noticing a lick of stutter in any games at all either.

Having said that, some games that showed significant uplifts at 4K output, using either High, Max or Optimised settings, and DLSS Quality through Performance modes.

• Space marine 2
• DOOM Eternal
• Spiderman Miles Morales
• Ratchet and Clank rift Apart
• Hell Blade Senua's Saga
• Crysis Remastered

 

[phanbuey]

Yeah if you heavily tune both chips then you have to wait a bit for the 5090 or 60 series to see a difference. By then we might have Nvidia N1/N2 with unified memory smashing everything.

 

[rv8000]

Your heavily tuned 12900K is surely a better example than stock which reviews would show, and I'd wager it's on (tuned?) DDR5 not DDR4 like my set and forget XMP DDR4 on the 5800X3D.

I can't really comment with certainty around 9800X3D vs heavily tuned 12900K, my point before was that I'm not noticing a lick of stutter in any games at all either.

Having said that, some games that showed significant uplifts at 4K output, using either High, Max or Optimised settings, and DLSS Quality through Performance modes.

• Space marine 2
• DOOM Eternal
• Spiderman Miles Morales
• Ratchet and Clank rift Apart
• Hell Blade Senua's Saga
• Crysis Remastered

POE2 is also known to have stuttering issues not explicitly attributed to a specific CPU, nor is poe a full released game as its still in early access.

 

[l1CappYl1]

I'm mostly playing poe2 but I've noticed it in pretty much every game. I'm running afterburner, fancontrol and hwinfo on the background, same stuff I've been using on every computer.

Have you tried without afterburner? I saw somewhere that it was the cause of stuttering on some x3d CPUs.

These chips scream for a delid and direct die water cooling. There's just no better way lol.

I finally got my hands on one and am planning to do just that. Just hope the delid process goes smoothly as I've never tried with a soldered CPU and am a bit worried something with crack.

 

[ShrimpBrime]

Have you tried without afterburner? I saw somewhere that it was the cause of stuttering on some x3d CPUs.

I finally got my hands on one and am planning to do just that. Just hope the delid process goes smoothly as I've never tried with a soldered CPU and am a bit worried something with crack.

Nah, the tool makes it pretty simple. Watch a couple videos, take your time follow procedures, and Bam! De-lidded! Then some core lapping.... so forth. Should be pretty sweet!

I suggest lapping the waterblock with the same surface you use to lap the cpu though. I have a really thick mirror from an old infrared alignment machine that I use for example.

 

[l1CappYl1]

Nah, the tool makes it pretty simple. Watch a couple videos, take your time follow procedures, and Bam! De-lidded! Then some core lapping.... so forth. Should be pretty sweet!

I suggest lapping the waterblock with the same surface you use to lap the cpu though. I have a really thick mirror from an old infrared alignment machine that I use for example.

I've done some lapping on the ihs previously, but never on the core itself. Seems pretty risky.
I have the direct die block from Thermal Grizzly, so I'll assume that it is already the correct thickness required.