RTX 4090 vs RTX 3090 | VR Performance Comparison | Tested in Microsoft Flight Simulator

well many of you have also here it is a performance comparison between the RTX 1390 and the RTX 1490 graphics card I recently purchased the RTX 1490 directly from Asus in the UK and it's their tough gaming model now these tests will not be exhaustive of course I've only had the card for a week so still lots of tests and settings to do so this video is in no way a settings guide but what I hope it will be is indicative in terms of what sort of performance you can expect in virtual reality in your VR headset with a RTX 1490 graphics card welcome to the Sim Hangar my name's Mark thanks for watching and let's get started as a VR user I guess I'm no different to anybody else I'm always looking for the next tweak or setting or something that will give me an extra frame or two and really satisfied with performance well in my personal opinion the leap in performance between the 3000 and 4000 series Nvidia graphics cards is significant it's more than what we've witnessed before and in my opinion VR users are likely to gain the most benefit from that leap I've tried to make the tests as real world as possible in fact I've probably gone a little overboard with the denser scenery area I could find which is Tokyo in Japan as well as a combination of default aircraft as well as third-party aircraft such as pmdg 737 we'll have a quick look at settings and then jump into the tests themselves but first I want to answer a couple of quick frequently asked questions firstly dlss3 is not available for VR it's 2D monitor mode only secondly you may see some settings within Microsoft flight simulator That You Don't See in Sim yourself and that is because I'm on Sim update 11 beta if you want to join the beta you can find out the details through the main Microsoft flight simulator website but there's a number of caveats to joining so make sure you read the details carefully and thirdly what is the maximum number of FPS frames per second that I can get on the Reverb G2 while the answer is 90. a VR headset is no different to a monitor the maximum number of frames that can be displayed on your monitor or your VR headset is governed by the refresh rate of your monitor or your VR headset in the case of the HP Reverb it's got two mode 60 and 90 90 is the default so 90 frames per second in that case would be the maximum number of frames that can be displayed onto my system settings I'm using a 10th gen 10900k Intel CPU processor 32 gigabyte ddr4 memory the EVGA RTX 3090 and the Asus tough gaming 1490 as mentioned previously as I'm using the HP Reverb I'm using Windows mixed reality my openxr tools for Windows mixed reality is set at render 100 motion reprojection is disabled I'm not using the openxr toolkit that has been disabled in the Nvidia control panel under manage 3D settings both power management and texture filtering quality are set to high performance under Windows game mode is off and Hardware accelerated GPU scheduling or Hanks is on for my 4090 tests and off for my 1390. here are the VR settings I've used for all the tests and simply change the anti-aliasing mode as needed for comparative purposes as well as swapping between DX 11 and dx12 Nvidia reflex low latency was on why wouldn't you and these settings reflect what I used for my 3090 as my set and forget settings assuming TAA was my default anti-aliasing method that is it would give me between 30 and 35 FPS in high density complex situations and were chosen because I don't want to keep coming to and fro my various settings and changing them depending on where I'm flying or what aircraft I'm flying could I up these settings with a 4090 I'm sure I could but that's a different subject and a different video for another time whenever I do one of these videos I inevitably get the comment it's not about the FPS it's about how smooth it feels and that's a true and accurate statement and is the most important factor in VR overall but for videos such as this the FPS is a fairly good indicator of performance and before we get into the tests I can state that the 4090 performance was considerably better not just in fps but in the absence or reduction in the number of microstutters pauses and so on but again this is not definitive it's for you to form your own opinion let's get on with the tests foreign in this first test the 4090 recorded a 60 Improvement in fps in this low-level ROM through Tokyo dx11 dlss quality mode whilst both cards at times recorded limitations by the GPU the 1490 was far less as it can process faster and this is borne out from the above graphs we can see the 1490 latency of 11.6 milliseconds compared to 22.9 milliseconds for the 3090 the 4090 remained constrained by the limitations of the main thread which is the CPU and as a result of Microsoft flight simulator being predominantly single core application let's now compare how dx12 fares foreign we're in a very dense scenery area right in the middle of Tokyo the increase in fps was about 15 or 38 percent the limitations of the 1490 were almost exclusively main thread or CPU whereas with the 3090 the limitations was a combination of the CPU and GPU what this indicates to some extent is that your CPU will have a marked bearing on the results that you have if you've got a 4090 if I had a 13th gen Intel CPU I would expect better results than my 10th genus providing let's do another test in dx12 same settings [Music] foreign the default a 320 this time overhead Tokyo over the complete flight the 4090 recording 55 FPS compared to an average of 42 for the 3090. that's about 31 the 4090 aren't performing the 3090 through share Grunt and we can see that with an average latency of around about 11 milliseconds compared to 20 21 milliseconds for the 3090. it was interesting to note that whilst the 4090 remained well ahead of the 3090 on Final Approach just before touchdown the FPS dropped significantly and the difference being fairly marginal I've used this example deliberately to highlight a couple of points I was aware from previous experience that approached to this handcrafted the Sobo airport tended to knock frames quite heavily this is hanida airport Romeo Juliet tanko Tango in my experience that has a bigger impact than JFK and the default Heathrow Airport when it comes to things such as shaders and textures your CPU has a lot of computation to do and if you're limited by your CPU processing power and single core processing you're going to get a drop in frame rates regardless of how powerful your GPU is nonetheless keeping this in perspective this is a flight sim a mid 40 FPS in VR well it's an amazing performance but as we've said before it's not just about frame rate it's about how it feels and the 4090 well it felt smoother they were less micro pauses [Music] [Music] as you've probably gathered that was pmdg 737 once again overhead Tokyo and on final two Romeo Juliet Tango Tango one of the most complex and demanding aircraft in one of the densest scenery areas in Microsoft flight simulator a real stress test and we took it one step further by using tea as the anti-aliasing mode and we can identify the additional stress placed on the system by the increase in latency for the GPU in both instances both were the 4090 one in dx12 and one in dx11 the 1390 results for the same flight were not surprising both being in the lower 30s for anyone interested in the route while we started in the air a low level circle around Tokyo itself then it turned directly overhead for an approach directly to the airport fairly quick flight for testing purposes around about 14 nautical miles I do need to mention on Final Approach on both instances I got one big pause one question inevitably I'll be asked is what happens if all the settings are on Ultra what sort of performance will I get then well I haven't tested that and why well putting all your settings on Ultra is self-defeating bear in mind we have a 2160 by 2160 per eye resolution in the HP Reverb G2 turning everything up to ultra will put more load on the system but not necessarily improve the visuals that I see perhaps those with the vojo era will gain some benefit more than others so having done all these tests what conclusions can we draw and we need to bear in mind that these tests have really been stress tests putting the systems under load as much as we could and not necessarily representative of your day-to-day flying chances are in day to day flying the 49 T results are going to be even better and my most important conclusion is the 4090 for me does not disappoint it's a beast of a card when flying in high density areas with complex aircraft we can expect an uplift between 30 and 40 percent I would suggest the average uplift is going to be between 45 and 55 for day-to-day flying and in light scenery areas with aircraft that are well optimized well a hundred percent certainly not out the question if you haven't done so already check out this video of the mb339 jet flying through New York FPS averaging around 60 and Topping 80 on occasions now I'm not sure if it's Sim update 11 or the 4090 but when using dlss2 mode under the quality setting the blurriness in the instruments and in the numbers has disappeared not quite sure why but I'm hoping it's some update 11 and they've got on top of this problem the next point I want to raise is dx11 or dx12 which is best while results May Vary system to system but for my system dx11 gives a better fps but my choice now is dx12 and the reason is that the frame rate count is far more stable dx11 frames tend to vary there's some big Peaks and some big troughs with dx12 we still get Peaks and troughs but not as extreme dx12 seems more stable and more consistent what I haven't done however is some extended long flights my longest flight with dx12 at the moment has been about an hour hour 20 min whether or not due to memory leakage performance declines over a period of time I don't know at this time and the last point I want to raise is one I've touched on already and that perhaps now more than ever your CPU and speed of memory can have a big impact on the results you experience with the 4090 ignoring for a moment these single core architecture of Microsoft flight simulator the GPU is no longer the bottleneck the bottleneck for your system now turns onto your system's ability to process the information faster so when deciding to plumb for the RTX 1490 graphics card or not there are other considerations to take into account there's no getting away from the fact that the 4090 is an expensive graphics card and in many instances will be hard to justify and will therefore fall into the Enthusiast category but for those committed to VR it does without question hold the promise of performance levels we haven't seen before and it does so to a level where getting an extra one or two frames really is not that important be that in dlss mode or be that in TAA mode the choice is yours so what are the best settings for the 1490 and Microsoft flight simulator well in Sim update 11 there's a couple of new options so there's a lot to test and things continue to change in Sim update 11. so the short answer is I don't know more tests required and I'll probably wait until summer update 11 has been out officially and then I'll follow up with a video but for now it's back to experimentation thank you very much for joining me today I hope you found it useful and informative stay well look after yourselves as always see you again soon and bye for now

Channel: SimHanger Flight Simulation

Views: 46,418

Keywords: msfs rtx 4090, rtx 4090 msfs, msfs 2020 rtx 4090, simhanger flight simulation, msfs dlss settings, msfs dlss vs taa, msfs dlss, rtx 4090, microsoft flight simulator, 3090 vs 4090 benchmark, 3090 vs 4090

Id: IXgXs3IfwjY

Length: 17min 4sec (1024 seconds)

Published: Wed Oct 26 2022