Jump to content

Reset XPDR

Members
  • Content Count

    2,308
  • Donations

    $0.00 
  • Joined

  • Last visited

Community Reputation

39 Neutral

About Reset XPDR

  • Rank
    Member - 2,000+

Profile Information

  • Gender
    Male

Flight Sim Profile

  • Commercial Member
    No
  • Online Flight Organization Membership
    VATSIM
  • Virtual Airlines
    Yes

About Me

  • About Me

Recent Profile Visitors

3,037 profile views
  1. My system is similar, except I have a 3080, and at similar settings (4K Ultra Preset DLSS performance) in a complex scenario (KLAX, PMDG 737) my GPU's frametime of 16 ms is around half my main thread frametime of 30 ms at that setting. Changing back to TAA still has me CPU limited and GPU frametime only goes up to 19ms. Either option gives me around 33 FPS and everywhere else simpler it is higher. Even my main thread frametime is 10 ms faster than yours, which is strange because the 10850K should be slightly slower than your 10900K. There is no way your 3090 should be delivering 3 times the frametime, which equates to 1/3 the performance, of my 3080. There is something wrong with your setup and I'd start by running a program like HwInfo to make sure your CPU and GPU are running at full speed when running MSFS.
  2. Because that is a synthetic benchmark. Find actual MSFS benchmarks, like the ones posted later in this thread, to show the relevance to using these processors in MSFS. Here's another two showing how the 5800X3D holds its own against the 12900K(S) in MSFS: https://www.youtube.com/watch?v=EOL2sDXK64A&t=208s, https://www.youtube.com/watch?v=ozM92v4KhIk&t=247s
  3. This one hands the win quite handsomely to the 5800X3D by a margin of about 20%+. In any case, I expect that from the OP's perspective the performance of the 12900K in MSFS should be moot as a 12900K, especially if paired with DDR5, in not a budget solution. The 5800X3D is a cracker of a budget CPU for MSFS and will easily met their stated requirements.
  4. To add to my response above, check out the following post which shows how a 5800X3D/6800XT combo is performing with the MSFS SU10 beta. At 1440p, DX12, Ultra Settings and LODs at max, the 1% low framerate is 68.3 FPS, more than double your desired outcome!
  5. If I had to choose between the two options, I'd pick the 5800X3D/6800XT as the GPU is still a solid choice for w1440p and the CPU is renowned for being a minimal stuttering MSFS powerhouse.
  6. A solid 22 FPS indicates that you are using motion reprojection at 1/4 refresh rate. The next step up is 30 FPS (1/3 refresh rate), so you will not see FPS increase to this until your settings allow averaging at above 30 FPS.
  7. As you say, it is hard to find MSFS benchmarks of this combo, but I found one site that had them for the https://www.gpucheck.com/game-gpu/microsoft-flight-simulator/amd-radeon-rx-6900-xt/amd-ryzen-7-5800x/ultra and https://www.gpucheck.com/game-gpu/microsoft-flight-simulator/amd-radeon-rx-6800-xt/amd-ryzen-7-5800x3d/ultra. While it doesn't show 1% low performance, they do show average performance at w1440p Ultra settings in MSFS to be just over 55 FPS for the both of them. Given this stated performance target, either option you present (5800X/6900XT or 5800X3D/6800XT) should achieve it through sheer average FPS being nearly double what your stated 1% low performance floor is and the fact that charts like below (albeit at 1080p) show 1% low performance with either CPU is no worse than 80% of average FPS. That is, if average FPS is 55 then 1% low should be no worse than 44 FPS, which is well above your stated floor. TL;DR - either option should do the job.
  8. In VR? In any case, MSFS gives us the luxury of telling you what is bottle-necking your current system, which in your case is the GPU, and by how much too. If you are significantly GPU limited and choose to upgrade the CPU, don't be surprised if performance is only marginally better, if not the same, because the GPU will still be the bottleneck. This would be a much better use case to benefit from upgrading your CPU.
  9. My system is pretty much always GPU limited in MSFS VR too. Of note is that when my system starts, I have turbo boost disabled (so 3.6GHz max instead of 4.8GHz) to keep it running cool and silent for windows desktop operations. When I run MSFS, I usually turn turbo boost back on, but sometimes I forget and for the most part it makes barely any difference in VR performance, like maybe 1 FPS at most. The only time it does make a difference is where the CPU is being taxed, so PMDG 737 at complex airport with lots of multiplayer, and that's when I notice if I haven't enabled turbo boost. As such, I would take heed of what MSFS is telling you. If you don't do the airliner at complex airports thing and otherwise putter around in GA, then a CPU upgrade is likely not going to benefit you much.
  10. I slightly undervolt my GPU and CPU, have a case with the same number of inlet as outlet fans (3 & 1 + 2 AIO), and set my CPU AIO fans to run around 1500 RPM when MSFS is running. This gives me 60-65C GPU, 50-60C CPU and a relatively silent system when running MSFS.
  11. Same thing happens to me. To deal with it, once I go into VR I try to stay in it. If I do go out of VR, I close down WMR and 2D performance returns, but trying to go back into VR is asking for trouble, so I don't.
  12. I'm using a G2, so perhaps you are referring to someone else. Nonetheless, I haven't seen this DLAA option before so I'll give it a try.
  13. I can get DLSS to look good if I boost my WMR render scaling by as much as DLSS downsamples it by, but then I end up with the same performance I was getting with TAA anyway so I'm not seeing the point of it. It's almost like Asobo forgot to apply the upscaling part of DLSS to get you back to your original render scale with how blurry it looks.
  14. I ran my i9 10850K cooled by a 240mm AIO at 5.2GHz, HT off and 1.35V for well over a year without major issues. Despite it not causing issues running MSFS, I have found it won't pass some stress tests now that it used to pass at those settings, so I have backed it down to 5.1GHz and 1.325V and it is once again solid, with CPU temp under load running MSFS around 50-60C. I also use Throttlestop to hotkey switch between two profiles, one at full speed (5.1GHz) and the other with turbo disabled (3.6GHz). Interestingly, in MSFS VR where I am mostly GPU limited, I get similar performance running my CPU at either speed. It is only when I load up a complex aircraft like the PMDG 737 at a major airport that the extra CPU speed makes a difference in performance, so for the most part I just run my CPU in non-turbo mode and just turn it on when I notice it needs it, just like the good ole days with the turbo switch on my 286-12!
×
×
  • Create New...