So here are the promised comparative videos with the 2080ti. I’d suggest opening the corresponding videos side by side in two browser windows, The flight are not identical so they one gets ahead of the other. I’d pause the one that gets ahead and wait for other to catch up and then hit play. Also I didn’t take the exact same taxi ways to the rwy every time. Alternatively, don’t bother and just take my word for it. The 2080ti improves performance over the 1080ti in an obvious and significant way. If I get $750 for my 1080ti the $450 or so difference will be the best value $450 for improved FS performance I have perhaps ever spent. I estimate that FPS increase is on average 50% plus so going from roughly 20fps to 30fps I am paying $45 per frame. I think that's pretty darn good. Some said that going from a 1080ti to 2080ti was not worth it if you played in 1080p. I disagree. The one caveat is that I do suspect that a very high CPU clock is probably required to really reap the benefit of a 2080ti in Prepar3D V4. My 5.3ghz on all six cores qualifies. As I suspected in the clear skies comparison where the CPU is the bottle neck there was little difference between the 1080ti and 2080ti. Who wants to fly a noon in clear skies all the time? But in all 3 other comparisons it looks to me that 2080ti wiped the floor with the 1080ti. New hardware means new optimization so I suspect I'll manage to optimize setting further. Final though is that I am very very happy with this purchase. You millage may very! I can’t wait to get Microsoft Flight Simulator onto this system.😁🤣😂

 

Edited December 14, 20196 yr by Avidean

Thanks for posting the results of your tests, Avidean.  The video card was not a huge player in FS9 or FSX, but P3D, X-Plane, and I expect the upcoming MSFS, do make better use of the video card.

Very interesting and well done Dave. Thanks for taking the time to record the videos and present this. I was surprised to see that even the 2080ti was nearly fully utilized in the overcast conditions. You must really have your settings cranked up. It does look like your system is better balanced now and the performance much improved. Glad you got a new card out of the deal after the hassle and stress you had to go through.

You can see my system specs below. My vintage 3770k is definitely the bottleneck but still very enjoyable to fly on my 4k screen as long as I keep my settings under control, i.e. medium and no dynamic anything. I am currently building an 8086 system so your presentation here was very informative for me. I was planning to use my 1080ti as long as I can stand it, which may not be long according to your videos.

Ted

As my final word on the subject I should point out on the 2080ti VR video it appears in the fpsVR window that the GPU is barely being used at all. Not so, I have always found that for some reason with fpsVR the GPU use doesn't register on the graph most of the time. You can see on the 1080ti VR video that it does register. Although for that one I had to restart the program or the order in which I started the programs to get it to work. I tried a few times to get it to register GPU usage on the graph for 2080ti VR video but I gave up since the GPU usage is also register on ASUS GPU Tweakll.

Flying in VR with the Samsung Odyssey+ has greatly improved. Although the GPU still ends up being the bottle neck. It has allowed me to go one notch up on the auto gen sliders and load in much of the eye candy that had to be dispensed with, with the 1080ti. I now have every shadow box ticked in P3D GUI and the Shadow Quality slider maxed. I also have all but sim object reflections ticked and all of that in overcast conditions while maintaining an acceptable FPS. Personally I can tolerate a lower Frame rate in VR than I think most people can. I find that 24fps in VR is about the same as a smooth 30fps on a screen. That would be the worst case scenario.

Anyway I can still get P3D to bring this system to its knee's. If I do the right thinks I can get 6fps. I am really looking forward to getting Microsoft Flight Simulator and seeing all the goodies at 90fps in VR🤣

 

By way of finishing the story I did end up selling the 1080ti for CAD$700. That's CAD$100 less than I paid for it about 1 year ago. So the used 2080ti (except I ended up with a new one thanks to the RMA😁) ended up costing me $500. Everything in my system is used or recycled except the Memory and the ssd's. I think its a proven way to beat the insane prices of first gen PC hardware. But you have to be careful.

Having recently built a replacement for the 6.5y/o 3930K/GTX Titan I'm temporarily using the GTX Titan until I decide when to pick up a better GPU.  But here's the thing:  the sliders that impact the GPU it would seem can easily undo RTX 2080Ti.  So it's been compelling to wait until Ampere 3080Ti or whatever it turns out to be called appears. 

I typically run complex scenarios w/ reflections turned on, shadow quality at low, MSAAx4 or x2, terrain/cloud shadows off, and dynamic lighting off, HDR/bloom on.  In lesser scenarios I may add short distance cloud shadows or dynamic lighting.  But here's the thing:  I can hardly see the difference visually between toggling these on or off.  Can you, without really looking for it? 

The new build sans new GPU, and going to 64bit P3D from v3.4, made a world of difference, pardon the pun.   Those pieces, including a new case, 9900K, Noctua cooler, mobo, 2x2Tb M.2 NVme drives, 32Gb 3200mHz ram, Win 10, a new PSU, DVD/RW, several case fans came in somewhere around $1500-$1600.  Clearly worth the $$ for the enhanced flying experience.

What do you think, is the payoff for another $1100+ worth it for your actual experience flying?  Not in running benchmarks, but in flying?  Do you really notice the visual differences the better GPU provides?  To cope w/ the old Titan I just am sure to keep things adjusted to keep load under about 92% always and visually, it looks rather stunning still.  3440x1440.

It's also a possibility to pick up an RTX 2080Ti now and then upgrade again when Ampere arrives, selling the 2080, which of course is why I am asking these questions!

I must be missing something cant see the GPU test numbers below.

  https://www.guru3d.com/news-story/rumor-nvidias-geforce-rtx-3000-video-cards-launch-delayed-to-september.html