RTX 2080Ti vs GTX 1080Ti: Is it worth it?

[funkyfranky]

 

 

No, I'm not gonna do this - this price policy is insane!

:drink:

:beer:

:drink:

 

 

• Situation: Free flight in the F-86F on different maps in areas with high object counts (large cities).
  
• Measurement: Replaying same track file with identical DCS settings. FPS saved to csv using FRAPS.
  
• DCS version: 2.5.4.29167
  
• NVIDIA driver: 419.67
  
• Oculus home: 1.37 (beta)
  
• Pixel density: 1.5
  
• Hardware:
  
  • Oculus Rift, ASW off
    
  • Asus GTX 1080Ti Strix OC
    
  • Asus RTX 2080Ti Strix OC
    

 

Caucasus: Batumi

(Click to enlarge)

 

NTTR: Las Vegas

(Click to enlarge)

 

Normandy: Caen

(Click to enlarge)

 

Persian Gulf: Dubai

(Click to enlarge)

 

Now, is it worth it? Dunno, your choice :)

[FragBum]

But.,, you should be able to crack 90FPS in Nevada and Caucasus away from built up areas I find the other maps to be more CPU resource hungry. ;)

 

Is it worth it? Yes, is it value for money, no. :D

 

CPU clock and RAM speed also help for me with Rift.

 

YMMV :thumbup:

[twistking]

Thanks for the comparison.

I'm planning to get into VR, but right now i'm still saving for a new system, so this is helpful.

 

Quick Question:

Why does the framerate seem to be capped at 45 so often, when AWS is off. Will the rift always cap to 45 (when it cant reach 90) even without AWS?

 

If so, why would one want to disable ASW? Is it just for testing purpose, or do you prefer it without ASW?

 

Thanks!

[dburne]

Thanks for the comparison.

I'm planning to get into VR, but right now i'm still saving for a new system, so this is helpful.

 

Quick Question:

Why does the framerate seem to be capped at 45 so often, when AWS is off. Will the rift always cap to 45 (when it cant reach 90) even without AWS?

 

If so, why would one want to disable ASW? Is it just for testing purpose, or do you prefer it without ASW?

 

Thanks!

 

I prefer without ASW, and no for me it does not cap it at 45. There is an option to set ASW off and cap it though, so maybe choosing the wrong option?

I use the Oculus Tray Tool to set mine off, CTRL +_Numpad 1 turns it off as well.

CTRL+Numpad 2 turns it off and caps apps at 45 Hz.

CTRL+Numpad 3 enables ASW and forces apps to 45 Hz.

CTRL+Numpad 4 enables ASW (default setting).

[funkyfranky]

Why does the framerate seem to be capped at 45 so often, when AWS is off. Will the rift always cap to 45 (when it cant reach 90) even without AWS?

Don't know exactly. Somehow the FPS like to snap to 45 FPS even with ASW off. And it was definitely off as one can see from FPS being between 45 and 90. That does not happen with ASW off. Also OTT shows ASW status off.

If so, why would one want to disable ASW? Is it just for testing purpose, or do you prefer it without ASW?

I just turned it off for testing purposes for the reason above, i.e. not having only 45 or 90 FPS. Usually I have it on as it makes things much smoother.

[Strong05]

What is your CPU clocked at?

[mhe]

Be prepared to do that all over again once the VR performance improvement drops that Wags spoke about!

[toutenglisse]

ASW off "bugs" since a long time (I'd say 1 year at least) with oculus rift, in sense you are stuck at 45fps many times, when you should see fps always varying from 0 to 90.

 

Funkyfranky that bug don't help you to make a clear comparison. What is strange is that 2080Ti don't show this behavior in NTTR and Caucasus map ?

[Krupi]

I would love to see these compared to the 2080 and 2070

[funkyfranky]

ASW off "bugs" since a long time (I'd say 1 year at least) with oculus rift, in sense you are stuck at 45fps many times, when you should see fps always varying from 0 to 90.

I did not know it's a bug but that's definitely what I observe for quite a while now.

Funkyfranky that bug don't help you to make a clear comparison. What is strange is that 2080Ti don't show this behavior in NTTR and Caucasus map ?

Well, also the 1080Ti shows values between 45 and 90. But is seems it only happens when it is well above 45.

 

So yeah, the comparison gives more a rough idea and the number should be taken with care.

[Harlikwin]

Weird, I'm still mostly locked at 45fps with my 2080ti with occasional dips lower than that. Similar settings as well, though I'm running 1.8PD. Then again I've done 0 optimizations as I just got the rig running.

[funkyfranky]

Weird, I'm still mostly locked at 45fps with my 2080ti with occasional dips lower than that. Similar settings as well, though I'm running 1.8PD. Then again I've done 0 optimizations as I just got the rig running.

You have ASW off? Because otherwise you will mostly see the 45 FPS in "normal" missions.

 

PD 1.8 compared to 1.5 is quite a bit more demanding already. Also bear in mind this is a free flight mission in the F-86. With more AI units around and say and a Hornet or Tomcat you also loose some FPS. Therefore, I think you are just fine.

[Harlikwin]

Well im hitting ctrl 1. Im not surenits actually going off tho.

[Cavemanhead]

Great data and well presented. Thanks for taking the time to do this... Interesting to see was the performance difference in the maps that ED may find very useful.

 

 

I missed the reason for the "switch" of 40+ FPS to double that multiple times by the 1080 Ti. What was holding it back?

[FragBum]

@Harlikwin

 

In NTTR away from Las Vegas you should get 90FPS happily on the ground in Gazelle, I run PD 1.2 and MSAA @ 4.

 

Caucasus depends where you are 45~90

 

PG good luck,

 

Normandy eats even more CPU usage.

 

What CPU and GPU usage are you seeing?

[hansangb]

 

No, I'm not gonna do this - this price policy is insane!

:drink:

:beer:

:drink:

 

 

• Situation: Free flight in the F-86F on different maps in areas with high object counts (large cities).
  
• Measurement: Replaying same track file with identical DCS settings. FPS saved to csv using FRAPS.
  
• DCS version: 2.5.4.29167
  
• NVIDIA driver: 419.67
  
• Oculus home: 1.37 (beta)
  
• Pixel density: 1.5
  
• Hardware:
  
  • Oculus Rift, ASW off
    
  • Asus GTX 1080Ti Strix OC
    
  • Asus RTX 2080Ti Strix OC
    

   

 

--actual slides removed--

 

Now, is it worth it? Dunno, your choice :)

 

 

Interesting. I guess my impressions are correct. I find NTTR to be smoother and always chalked it up to desert motif. I figured Persian map is the most "optimized" but I guess not. Thanks for taking the time.

[funkyfranky]

I figured Persian map is the most "optimized" but I guess not.

Well, I would not draw too many conclusions between the maps to be honest. For that the situations are too different. For example, on PG the track was recorded directly over Dubai. So the object count is probably way higher than on the other maps even NTTR over Vegas.

This comparison was meant more for the two graphics cards under otherwise identical conditions (per map).

[Harlikwin]

@Harlikwin

 

In NTTR away from Las Vegas you should get 90FPS happily on the ground in Gazelle, I run PD 1.2 and MSAA @ 4.

 

Caucasus depends where you are 45~90

 

PG good luck,

 

Normandy eats even more CPU usage.

 

What CPU and GPU usage are you seeing?

 

I see 90 on the load screen and then when I look at the sky :). I'm running 1.8/2x msaa, I might just turn it down, but it seems really smooth with no stuttering at 45 so it doesn't really bother me. I haven't had time to really do any optimizations or look real closely at stuff yet, the little time I've had I wanted to "play".

 

I haven't looked really closely at CPU, but at a guess its hammering one 9700k core. I had a weird moment when I realized that my turbo boost was set to 4.6 and not higher which was weird. I bumped it to 4.8. But I haven't systematically gone through it to figure it out. On the plus side temps are fairly mild on the CPU ~60C.

 

GPU wise, I'm trying to not hammer the card too hard so I'm seeing like 70-80% utilization, but sometimes its way lower than that. I've noticed the card is sometimes power limiting. And temps are largely in the 70's.

[FragBum]

For me I find PD of 1.2 and MSAA at 4 to look okay but the lower PD will allow you to keep a good solid 45FPS. Especially in the PG map, that map is a CPU/GPU resource hog.

 

The other things I do is pretty much disable all the nVidia processing as well I disable effectively Rifts own anti-aliasing this frees up some CPU and GPU resources.

 

Just for giggles I have a free flight mission in NTTR starts at Tonopah test range and in Gazelle that is a solid 90FPS anywhere which is good to test and tune settings.

 

Alas shiny new CPU and all, and yet as far as I can see the CPU is the bottleneck in performance.

[twistking]

The other things I do is pretty much disable all the nVidia processing as well I disable effectively Rifts own anti-aliasing this frees up some CPU and GPU resources.

 

 

What kind of AA does the Rift have on it's own and how do you disable it?

I set up a Rift with DCS on a friends pc and we could not find any AA/PD settings with the Rift itself.

[dburne]

What kind of AA does the Rift have on it's own and how do you disable it?

I set up a Rift with DCS on a friends pc and we could not find any AA/PD settings with the Rift itself.

 

Not really AA, but one can SS with it using Oculus Tray Tool.

[twistking]

Not really AA, but one can SS with it using Oculus Tray Tool.

ok. thanks for the answer!

[eaglecash867]

You can also disable Rift's own AA through Oculus Home. Click on the settings icon on the Oculus Home control panel, then select Graphics from the settings menu. AA will be there to turn on and off.

[Harlikwin]

Yeah it definitely makes a difference in performance. Based off my old-n-busted experiences.