Jump to content

VR Hardware Benchmark late 2020 - Ryzen 2000/5000 - 1080ti vs 3080 vs 6800XT


Recommended Posts

Posted

I have Ryzen 5900X, Asus Crosshair VII X470 motherboard, 32 Gb of DDR 4 3600 Mhz Cl 15 memory, 750 W Evga Supernova PSU. DCS is isntalled on NVME drive.

 

I dont know anything about the other guy. Merely observed his thread and saw the same situation.

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted
10 minutes ago, Hiromachi said:

Hmm, but that changing usage sometimes leads to FPS going below 45 and re projection point which in turn reduces smoothness.

I've noticed another user also reported that:

 Also, my 3080 whenever I did not have issues, was maintaining stable 85+ % usage and did not drop frames in such manner.

Thanks! Interesting... In that case I also have to investigate. My GPU is jumping around between 30 and 99% within seconds in fpsvr. 

11 minutes ago, JayRoc said:

That's not true. 

Modern CPU clock down and throttle the power that goes through a CPU is they're getting too hot. The usage of the CPU will be 100% if it is fully used no matter if it clocks slower or faster. 

 

The usage of the CPU doesn't really matter because its mainly 1 thread of your CPU that is used to full extend from DCS which limits your frames.

 

@Hiromachiwhat hardware do you use? Did the other guy use DDU to deinstall the drivers aswell?

 

Is that typo or are you talking about CPU? Because we talked about GPU usage. 

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Posted

https://rog.asus.com/motherboards/rog-crosshair/rog-crosshair-vii-hero-model/helpdesk_bios/ Version 4007. Doesnt look like its a beta. And it would be fine if it was only my issue, but considering someone else reported this as well ?

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted

Since we have 3 different PCs with absolute no problems, it would make sense to start looking for differences from our systems to yours and maybe try to find something your and the other guys system have in common to pin point the cause. 

 

Something like a Ryzen 5000 on a x470 might be it. 

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Posted
11 minutes ago, Hiromachi said:

https://rog.asus.com/motherboards/rog-crosshair/rog-crosshair-vii-hero-model/helpdesk_bios/ Version 4007. Doesnt look like its a beta. And it would be fine if it was only my issue, but considering someone else reported this as well ?

So, just to clarify, you're dropping below 45fps in the benchmark test? Because even with the described drops between 30 - 90% GPU usage I never dropped below 45fps in that test.

And does your Gpu boost to the given boost clock and then drops down or does it even not reach the boost clock? 

 

Maybe test also with hwinfo and have a look at your maximum and average clock speed. 

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Posted

I wonder this myself. As said, I used the exact files linked in first post, following instructions in those files (i.e. when to start and stop benching).

 

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted

benchmark v1 complex mission.miz

 

This is my excact file. You could try this one.

Questions comming to my mind:

 

- any special beta version due to damage model testing?

- anything else that could differ regarding DCS?

- SS Resolutions the same when swapping cards?

 

(You could update your signature btw)

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Posted (edited)

No special beta, its just label under my name. I use same open beta as everyone else 🙂 
Dont think so, same modules and usual settings (especially with VR preset). Resolution in SteamVR with SS is 100 % - default one. I did not lower it like Bignewy recommends in his Reverb setup tutorial. 

I could, but Im waiting for forum to be set in stone. Had one in previous forum but all stopped working.

 

EDIT: So I repeated test. To make sure time is accurate, I paused game the moment message to start benchmark appeared and started benchmark run and then stopped benchmark when message to end benchmark was about to fade out. Still, 2.9 min.
EsioSy.png

To be fair its not clock fluctuation that matters but usage fluctuation. GPU reaches its designed clock speed and even goes as high as 2600 Mhz in some places. 

And yes I'm dropping at this very benchmark with mission provided by JayRoc below 45 FPS towards its end when overflying six small fires. Than FPS fluctuates between 40 and 50 and usage fluctuates between 20 % and 90 %.

Edited by Hiromachi

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted

I've enabled SAM. I only see a minuscule improvement:

94leUK.png

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted (edited)

We didnt pause the game, we startet the mission an then activated the autopilot.

After that we Alt+Tab out of the game and pushed the reset button once the start and end appeard.

 

The usage of the GPU is way different. (Bottom graph)

Do you see this behaviour in other games aswell?

Not sure if it was mentioned here: did you delete the shader cache?

 

image.png

Edited by JayRoc

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Posted (edited)

Fxo, metashaders and Radeon driver shader cache. Yes. 

 

I also have issues with reprojection. I see text warping in G2 and reprojection does not hold to 45 FPS. If GPU can push more than it goes up to 60+ but game is actually less smooth if it was just sticking to 45 FPS. 

 

Sidenote: I tried game on flat screen (4k, with all blows and whistles cranked to max and sharpening at 100 % in Adrenalin). GPU usage was at 95-99% at all times, GPU clock was constant at 2650 Mhz and FPS was around 90 - 110. The GPU usage fluctuations in VR are pretty much similar to what I had with Nvidia since few weeks. Since its VR exclusive issue I'd point my finger at SteamVR. 

Edited by Hiromachi

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted (edited)

image.png

image.png

image.png

image(1).pngThis looks great, as long as somebody plays low fi models in SP.

Throwing in a high fidelity model, and everything goes away.

Similar low level passes over Persian Gulf produce:

image3.pngimage2.png

Edited by ags
Pictures have been eaten

VR, i9, Winwing, TM and VKB.

 

2, 4, ban, ban, 2

Posted
2 hours ago, Hiromachi said:

Fxo, metashaders and Radeon driver shader cache. Yes. 

 

I also have issues with reprojection. I see text warping in G2 and reprojection does not hold to 45 FPS. If GPU can push more than it goes up to 60+ but game is actually less smooth if it was just sticking to 45 FPS. 

 

Sidenote: I tried game on flat screen (4k, with all blows and whistles cranked to max and sharpening at 100 % in Adrenalin). GPU usage was at 95-99% at all times, GPU clock was constant at 2650 Mhz and FPS was around 90 - 110. The GPU usage fluctuations in VR are pretty much similar to what I had with Nvidia since few weeks. Since its VR exclusive issue I'd point my finger at SteamVR. 

 

 

I got a much smoother gameplay with reprojection off. But it was set on in the benchmarks.

Do you run the steam or the standalone version? SteamVR beta? SteamVR for WMR beta? any other VR headsets(software)?

Which Windows Version? 20H2? Checked Microsoft Store for a WMR update?

1 hour ago, ags said:

great, as long as somebody plays low fi models in SP.

Throwing in a high fidelity model, and everything goes away.

Similar low level passes over Persian Gulf produce:

 

well...its a highly optimised game as we all know ;D

Online i stick most of the times with 45fps with my F16...

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Posted
10 hours ago, Hiromachi said:

Fxo, metashaders and Radeon driver shader cache. Yes. 

 

I also have issues with reprojection. I see text warping in G2 and reprojection does not hold to 45 FPS. If GPU can push more than it goes up to 60+ but game is actually less smooth if it was just sticking to 45 FPS. 

 

Sidenote: I tried game on flat screen (4k, with all blows and whistles cranked to max and sharpening at 100 % in Adrenalin). GPU usage was at 95-99% at all times, GPU clock was constant at 2650 Mhz and FPS was around 90 - 110. The GPU usage fluctuations in VR are pretty much similar to what I had with Nvidia since few weeks. Since its VR exclusive issue I'd point my finger at SteamVR. 

 

This! 

 

I was wondering the same. I am not able to hold steady 45fps with reprojection on some scenes with fps then dropping down to 30. Although I have around 15ms GPU frametimes. The same scene with reprojection off I manage to get 50 - 60 fps. I guess there is something wrong with SteamVR and the new AMD cards. 

 

I'll be back tomorrow on my rig and will try a fresh install of the software. 

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Posted (edited)
11 hours ago, ags said:

image.png

image.png

image.png

image(1).pngThis looks great, as long as somebody plays low fi models in SP.

Throwing in a high fidelity model, and everything goes away.

Similar low level passes over Persian Gulf produce:

image3.pngimage2.png

 

EDIT 2: I wrote a lot of bollocks and deleted it! 

 

So it looks like, during lofi SP missions we see a slight advantage for the 6900XT while the table turns with the last two graphs. But both cards seem to struggle with the discribed scenario (low pass). 

Edited by exil

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Posted
12 hours ago, ags said:

image.png

image.png

image.png

image(1).pngThis looks great, as long as somebody plays low fi models in SP.

Throwing in a high fidelity model, and everything goes away.

Similar low level passes over Persian Gulf produce:

image3.pngimage2.png

 

Mhh, the last two graphs doesn't even look like the same mission or scenario? 

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Posted
2 hours ago, exil said:

Mhh, the last two graphs doesn't even look like the same mission or scenario? 

Both are free flight, F18, PG, low level one way, reverse, low level the other, but driven by hand.

 

VR, i9, Winwing, TM and VKB.

 

2, 4, ban, ban, 2

Posted
13 hours ago, JayRoc said:

 

I got a much smoother gameplay with reprojection off. But it was set on in the benchmarks.

Do you run the steam or the standalone version? SteamVR beta? SteamVR for WMR beta? any other VR headsets(software)?

Which Windows Version? 20H2? Checked Microsoft Store for a WMR update?

To be fair all my VR problems (had RTX 3080 than) started in late Nov / early Dec. I wouldnt even bother with 6800XT or anything if not for massive stutters with 3080, indicated here: 

 

Anyway, to answer your questions.

1. Standalone version,

2. SteamVr non-beta, version 1.15.16 with hotfix,

3. SteamVR for WMR beta,

4. No other headsets,

5. Windows 10 Pro 20H2 compilation 19042.685,

6. WMR is up to date.

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted

I know future proofing is usually futile but with DCS developers talking about DLSS and raytracing coming to DCS in the future wouldn't a NVIDIA card be the safer bet if you're planning on buying a new card?  Just going on benchmark results with RTX and DLSS.   Not sure how Vulkan affects these things. 

Posted

Since when did ED claim to be using DLSS and Nvidia Raytracing ? Those are proprietary technologies and so far DCS never supported anything like that. Since DCS is going to change API to Vulkan, its likely that we will get some day Khronos open ray tracing stuff as well. And OpenXR support. 

 

Anyway, I quickly swapped back to 3080. My stutters are back but they can be cured with simple alt+tab. However reprojection works flawlessly on RTX 3080 and it holds those 45 FPS much better than 6800XT. And even going below / above it does not affect smoothness. Re-projection must have some issues with AMD.

Second thing is that text warping I had in DCS settings / controls with 6800XT is also gone. 

AMD Ryzen 5900X @ 4.95 Ghz / Asus Crosshair VII X470 / 32 GB DDR4 3600 Mhz Cl16 / Radeon 6800XT / Samsung 960 EVO M.2 SSD / Creative SoundBlaster AE-9 / HP Reverb G2 / VIRPIL T-50CM /
Thrustmaster TPR Pendular Rudder Pedals / Audio Technica ATH-MSR7

Posted

Ray tracing is not proprietary technology, the 3000 series is Nvidias second gen ray tracing whilst the 6000 series is AMDs first attempt.

 

Personally I think it will be a few generations before even the hybrid version of ray tracing we currently have won't be a huge hit to your FPS. 

Windows 10 Pro | ASUS RANGER VIII | i5 6600K @ 4.6GHz| MSI RTX 2060 SUPER | 32GB RAM | Corsair H100i | Corsair Carbide 540 | HP Reverb G2 | MFG crosswind Pedals | Custom Spitfire Cockpit

Project IX Cockpit

Posted
10 hours ago, Hiromachi said:

Since when did ED claim to be using DLSS and Nvidia Raytracing ? Those are proprietary technologies and so far DCS never supported anything like that. Since DCS is going to change API to Vulkan, its likely that we will get some day Khronos open ray tracing stuff as well. And OpenXR support. 

 

Anyway, I quickly swapped back to 3080. My stutters are back but they can be cured with simple alt+tab. However reprojection works flawlessly on RTX 3080 and it holds those 45 FPS much better than 6800XT. And even going below / above it does not affect smoothness. Re-projection must have some issues with AMD.

Second thing is that text warping I had in DCS settings / controls with 6800XT is also gone. 

I'm pretty sure that Nick Grey said in this interview that they are looking into DLSS & raytracing

 

 

 

 

And if it's not this then Simon mentioned it here:

 

 

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...