Caporegime
Yeah i'm going to run it now...Be interested to see your results. Just wondering how RDNA 3 improves things with the basic RT
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Yeah i'm going to run it now...Be interested to see your results. Just wondering how RDNA 3 improves things with the basic RT
I don't think it's anything worth tweaking extra over, already at good performance, any medium differences don't really mean anything in the grand scheme of things tbh beyond idle curiosity.I thought the 7800X3D was only better in niche titles that made use of the vwotsit, like Factorio, some MMO's etc. I'll change a bit in bios and see what happens.
And yes same. I've left discord and all those other gubbins open. I'm not a competitive benchmarker lol. Doubt it would make much difference anyway
I don't think it's anything worth tweaking extra over, already at good performance, any medium differences don't really mean anything in the grand scheme of things tbh beyond idle curiosity.
No it's the fastest gaming cpu period but if a game doesn't use the vcache well it will fall down but heres a look at the 7800x3d v the 14900k. At 4k though most top cpu's will run or should run similar performance. At 1080p/1440p is more where it will matter as you can see in the results.I thought the 7800X3D was only better in niche titles that made use of the vwotsit, like Factorio, some MMO's etc. I'll change a bit in bios and see what happens.
And yes same. I've left discord and all those other gubbins open. I'm not a competitive benchmarker lol. Doubt it would make much difference anyway
Weird, I got the benchmark tool this morning so unless they updated it shortly after then not sure what to add to that as it was definitely showing an older version for both FG and DLSS upscaling, Ray reconstruction (dlssd) was 3.7.10 though.I'll try the DLL files once I work out where / how to do it, just for science.
Yeah no idea what I'm doing, I've updated dlss.dll to 3.7.20. @mrk you said yours was 3.1.30 but mine was 3.7.10.
dlssd is current version a 3.7.10
dlssg says latest version is also 3.7.10 (techpowerup)
Weird, I got the benchmark tool this morning so unless they updated it shortly after then not sure what to add to that as it was definitely showing an older version for both FG and DLSS upscaling, Ray reconstruction (dlssd) was 3.7.10 though.
Oh no the version 3.7.20 is only for super resolution whihc I updated, I also updated the frame gen dll as that was not the latest. ONly dlssd was up to date as both frame gen and dlssd are 3.7.10 as the latest.
Settings / Change from first run / run above | Average | Max / Min | Low 5th | Total VRAM |
GPU 100% PL - +100/200 (Core/Mem) NO RAYTRACING | 114 | 117 / 105 | 109 | 9.8 |
GPU 80% - +2865 / 10501 | 96 | 114 / 79 | 85 | 11.9 |
GPU 115% - +2865 / 10701 | 97 | 114 / 73 | 86 | 11.8 |
GPU 100% - 2775 / 10501 EXPO enabled | 95 | 113 / 81 | 84 | 10.5 |
GPU 100% - 2865 / 10701 EXPO enabled | 96 | 115 / 81 | 85 | 10.6 |
GPU 100% - 2865 / 10701 EXPO enabled NVIDIA driver update 560.81 (from 70) | 93 | 114 / 79 | 83 | 10 |
GPU 100% - 2865 / 10701 EXPO enabled - CPU All Core Negative -15 | 95 | 113 / 82 | 85 | 10.2 |
" " Updated NVIDIA .dll to latest versions | 96 | 116 / 82 | 86 | 10.5 |
"" Closed all background apps | 99 | 118 / 84 | 88 | 9.2 |
"" Set 'prefer max performance' on NVCP power setting | 99 | 117 / 85 | 88 | 9.2 |
"" RAYTRACING OFF | 122 | 138 / 37 | 111 | 7.6 |
The game's benchmark is accounting total VRAM use inc the OS/BG apps, not the dedicated game only VRAM use.Why do you think there are weird differences in VRAM usage? For example, on the RT Off runs at 4k
Yours =7
Mine = 7.6 (last run on table).
Yet on the first run I did, the usage was 9.8 (RT was on in the settings but wasn't enabled for benchmark as hadn't restarted).
The game's benchmark is accounting total VRAM use inc the OS/BG apps, not the dedicated game only VRAM use.
The only way to get dedicated game only VRAM use is via RTSS showing the process-VRAM.
RX 7800 XT
1440P
Super Resolution: 100
Frame Gen: Off
Cinematic
RT Very High: 6
RT Medium: 12
RT Low: 13
Without RT: 39
And finally i copied @mrk
Super Resolution: 75
Frame Gen: On
Cinematic
FPS:80
Any requests?
Haven’t tried cinematic and RT will try tomorrow - on par with my undervolted 3080 12gb version but falls approx 7% behind when it’s maxed overclocked.. game is tough to run!RX 7800 XT
1440P
Super Resolution: 100
Frame Gen: Off
Cinematic
RT Very High: 6
RT Medium: 12
RT Low: 13
Without RT: 39
And finally i copied @mrk
Super Resolution: 75
Frame Gen: On
Cinematic
FPS:80
Any requests?
Haven’t tried cinematic and RT will try tomorrow - on par with my undervolted 3080 12gb version but falls approx 7% behind when it’s maxed overclocked.. game is tough to run!
Right so with the 4090 at stock:
Basically exactly the same lmao, so looks like the 12700KF/K is better at UE5 than a 7800X3D
Edit&* For reference, I haven't closed anything down, both Firefox windows are open in the BG, multiple media tabs (20+) and all BG apps remain open too. I just load games and play so no "game mode" or any other tweaking Windows.