Gunnars Driver

Good work! I hadn’nt the time to read the test before. I will though ???? What I seen so far with the 3090 it a lot faster than the 2080ti, but not fast enough to produce better frame times than my 9900KS when using the Reverb. With my 8KX, the 3090 frame times is in the proportion of 2-3x the CPU. This might change a bit when I tune DCS etc for the 8KX but I do not really see that the GPU frame time come even close to the CPU. My conclusion is that ypu need a really beefy GPU but can do a little less CPU with the 8KX.

I dont know what Pimax they used, but when I do DCS with my 8KX, the CPU has a walk in the park, and the 3090 really work to make up the cost of it.

I have the Reverb Pro(”G1) and while I did not have any troubles with the 2080ti, tuned for smooth ride at 45 fps reprojection I see a huge increase in performance with the 3090. My settings was tined with the 2080 ti so I more or less never see any stuttet( I dont like gaming with a non smooth experience). For me, the 2080ti was always the limiting factor and the 3090 made a quite big difference and now the performance gap is less between the CPU(9900KS@5.2) and the GPU. Still almost always higher frame times for the GPU. Havent tuned the settings to max for the Reverb, as I did get the 3090 mostly for pimax 8KX where the 3090 is indeed needed in DCS. I would say that you should be able to get it running well with the 2080ti + the Reverb. There shouldnt be any real difference in hardware demands between G1 and G2 as the use the same resolution. So my conclusion is that you should be able to get it running well with the 2080ti, if all computer hardware/software is well tuned. If tou have the money for the 3090 it will be a big lift and you should be able to use higher settings for a better picture/ gaming experience. But with the weong settings Im sure it is possible to bog down the 3090 as well. Also, many here report that they are CPU bond. I havent seen this for my self, but one poster recently deleted mods and after that his computer delivered the about same frame times as I have. If you by some reason is highly CPU bond, you might not see that big improvement with the 3090. In my case the CPU frame times decreased slightly so it might help in that case also.

The whole setup have to be checked. As for the oculus service: as dburne says. But I have stopped the service manually and sometimes forgot it without noticing any performance drop. Of course no unnessesary programs shoudl be running but it isnt oculus that make one end up with 11 fps. The PD and SS is only one part of the solution. Most of the other settings in DCS can be use to turn a immersive experience into a slide show, even with a 3090. So, all DCS settings need to be tuned correct to make DCS run nice. The initial goal should be to get the game run smooth, and first after it runs smooth the settings should be increased to gain in picture quality. Here’s some posts I made ablout the settings: https://forums.eagle.ru/forum/english/dcs-world-topics/input-and-output/virtual-reality/7134520-hp-reverb-g1-issues?p=7135688#post7135688 https://forums.eagle.ru/forum/english/dcs-world-topics/input-and-output/virtual-reality/7134520-hp-reverb-g1-issues?p=7139808#post7139808 https://forums.eagle.ru/forum/english/dcs-world-topics/input-and-output/virtual-reality/7134520-hp-reverb-g1-issues?p=7139821#post7139821 https://forums.eagle.ru/forum/english/dcs-world-topics/input-and-output/virtual-reality/7143652-3090-first-impressions?p=7147054#post7147054 fpsvr is invaluable(but is only 3 or 4 usd on steam) when tuning the settings.

I ordered the 8KX bundle 1october and the pimax homesite said ”delivery in 7 working days”. The 8KX was sent within the 7 Working days but they still havent sent the base stations and controllers. Soon two months and I can not really enjoy the 8KX without tracking. starting to be a bit pissed on Pimax for this.

This is exactly correlating with my experience with my 2080ti, and this was also the reason I had to tune the settings lower to get rid of stutter. my 2080ti settings was limited ( in DCS only) by the VRAM. We know 30X0 should be better to manage VRAM so 10Gb will probably worl better than 11 om the 2080ti.

Just got a link from a friend about MSI Afterburner beta being able to show VRAM usage for each process: I havent had the time to look it trough

I guess people game too different to find a solution that fits all. I have yet to find any missions that significantly put the CPU frametimes above the GPU's. Yes it got closer by changing the 2080ti to a 3090 but still... This might be due to the fact that the missions I play with my friends doesnt have the stuff that increases CPU time to red values. And, I dont need to find one of these really. We can use (for example) the mission in this thread for benchmarking: https://forums.eagle.ru/forum/english/dcs-world-topics/input-and-output/virtual-reality/289741-gryz-s-benchmark-results but for tuning I guess one should use their own typical mission to make the tuning as adopted to their own gaming style as possible.

I am a bit split on the Pimax. FOV is great. Performance for me with 3090 is good enough. Assetto corsa is magic with fantastic feeling of speed(as its also is in DCS at low level.). I havent got the base stations I ordered at the same time as the 8KX(thats really Bad from Pimax) and DCS is a bit off to play in 3DOF. So I have done a few testing flights but not that much yet, maybe a couple of hours. Need the base stations :mad: What I didnt know when ordering the 8KX was that Pimax use lenses that need ones eye to focus quite narrow to see sharp. All other optical equipment in the world is made to use ones "far distance" seight. The 8KX focal distance is somwhere between 0,5 and 1m, I think. If you are near seighted or young enough to see sharp at 0.5m you can handöe this. At least for a while. If you can se sharp at 0.5m but the eyes struggle to bend the lens you will get eyestrain after a while. A lot of people is writning about eyestrain and I think a very big portion of they actually have problem keeping the eye sharp at this short focal distance. On Pimax own forum one can read that this have been known by some Pimax users and they have tried to get Pimax to tell them the exakt focal distance, but Pimax do not answer tihs. I did write a post about needing to now but I got the answer that "we have asked but Pimax wont tell us". Most people in forums are believing that it is the IPD thats the problem and they continue to tune IPD. So this means, as Im 53 and have gotten presbiopia(aging eyes with a lens getting rigider) I cannot see sharp in the 8KX. I am a Airforce pilot with still 20/20 or better seight at distance but I dont se sharp at distance closer than 1m. I also have a civilian ATPL with medical class 1 with the limitation near visual lenses(means I need to bring reading glasses but no glasses for the far distance seight). I see the picture in my Reverb very sharp and also the Oculus Rift. But I can not see the picture in 8KX sharp at all. I have a initial solution by using reading glasses and I plan to go to the optician and get perscription glasses that make me see sharp at the pimax distance(but I would need to know this to get it right). Glasses somtimes cause shine from the light inside, and this is not ideal but I think I can live with this. I feel that colors are much better than Reverb G1, but I think reverb is a little sharper in the sweet spot. The standard speakers built in into the 8KX is the worst crap ever seen. Supposed to send the sound from the speaker above the ear but its mostly not heard at all. I recommend anyone that plan to get 8KX (or probably any Pimax) to print a paper with small size letters and hold it at 0.5m. If you cannot see that sharp you might run into problems. If you are near seighted = congratiolations, get pimax. I had two friends on visit to make sure I didnt understand this wrong, both nearseighted and use glasses for this. Both was very happy because they vould see sharp in 8KX but they can not in my Reverb.

I will, had to little time left. Be back tomorrow.

Yes, I have no mods. Have never cleared the shaders either. my friends tested a new mod that sorts out data outside a ”VR-cone”. They say it made a big positive change. Dont remember the name of this mod but I can get back. Cannot be used with pimax though as it has to wide FOV.

Heres the same mission starring at the tower from the initial position in the F14: Remember I'm at SS100% if you didnt notice this. Might wanna read my the #16 post https://forums.eagle.ru/forum/englis...86#post7145186 I didnt notice the dropped frames until posting the pic, I never see dropped frames otherwise. There was no spikes this time.

So I found the mission acording to the picture below ( F-14 instant action bottom mission). Suggest we find a common solution to check tings. I use Open Beta, I dont know if theres any big differense to the official releasde version. This is the fpsvr text report: The 99.9 percentile does happen when I alt tab to start and stop the logging so If not doing this they isnt present. Also, Alt+Enter to make the screen fullsize(lowers framtimes) also causes frame spikes so I didnt use full screen window. A started the mission, activated fpsvr logging and then taxied to a catapult. Didnt look at the tower this time, but it only causes the frame times to increase slightly( 13ms maybe).

In my own mission, select a engine running F-18 on the supercarrier I get about 8ms CPU fram times. Looking at the tower increases the frame time to slightly above 11.1ms so FPS falls to 45 but rarely. As the fram time is this low, I can tune down the CPU to 5.0 without any loss at this position. I also did as supmua suggested, but I didnt tune down the settings tgo the very low VR-preset. Id never fly with those pretty low settings, not even when I hade a 1080. So pressing F2 + zooming away so I can do the 360 around my F-18 i still had about 8ms CPU time, during the complete 360 turn I had one red spike on fpsvr on the CPU, so short that I didnt notice any lag or stutter. All other times except that was about 8ms and in the green band for 90fps. I don know if theres any standard mission with the SC where we can use the same mission ? When I fly with my friends I dont notice any stutter so the frame time dont go above 22ms. When I tune DCS settings I tune to have some 25% margin to 22.2ms (that was with 2080ti+reverb...havent tuned it now, and I dont think Ill tune for the Reverb with the 3090. Tuning will be for the 8KX, thats the absolute main reason I did get the 3090.)

Is this a f-18 or F-14 instant action? What I used was a multiplayer mission that I use to host. Maybe a difference in loads due to other aircrafts/AI ?

What would the purpose be iof doing that?

Do you press A for autopilot ? Should activate two lights on the left lower side of the instrument panel if I dont remeber it wrong.

Yes, I do not feel that the supercarrier is that taxing. Gunslingers question was specifically about the SC. In my case, I didnt have fpsvr running but Id say from what Ive seen in lot of instances that the limiting factor probably wasnt the CPU and that I guess I can downclock to 5Ghz without loosing my 90 fps. ( I really never see my CPU limit, the frame times is always lower than the GPU. Changing from 2080ti to 3090 reduced the gap but the GPU still produce higher frame times. Even at this home cocked mission with me as a server and quite some AI in air, on ground and at sea) Also, I havent really tuned it fot 90fps on the SC, I really just happend to use the reverb on a test setting för my 8KX. For me the Reverb(= WMR) works fine with reprojection. No stutter at 45fps and no bad reprojection causing ghosting, this provided that I have sufficient margin to the frame times limit, and that the VRAM isnt full or close to full. Havent filled the 3090 VRAM but the 2080ti could be filled by high settings, causing stutter. For settings that make the F-18 smooth in flight but as high graphics as possible with 45 fps reprojection the F-14 most often isnt smooth. If I do a quick roll its obvious tha the horizon line stutters during the roll. I dont know why but it is like this for me, the same with AJS37.

So I added the F-14 to the mission. (The mission contains quite a few AI asset, both in air, on ground and at sea.). I didnt see any difference with the F-14 on the SC. Still 90 fps, unless starring towards the bridge, with the settings from above. The latest update broke the flight controls, just added the axis to be able to taxi. BTW= I use DCS open beta.

Absolutely! I see you and I agree on this, but for the record when the allocated term comes up: There has been a lot of people saying that the allocated VRAM isnt the same as the used one. By this they mean that when we read 17Gb or so, the real usage is below 10(or 11 or 8Gb). The main thinking is that the GPU allocates memory on the VRAM but does not at all use that much. What they do is to try convince them self that 10Gb(or 11 or 8) is enough. I say that as the VRAM is exlusive for the GPU there is no need to allocate memory. (Source: https://devblogs.microsoft.com/directx/gpus-in-the-task-manager/ ) The conclusion is that the task manager VRAM value is correct. Also, the "apps" reports is said to differ, but I find the delkta small and at least fpsvr seems to report slightly lower than the task manager. When we read 11 or 17 och 19Gb usage of VRAM, its not that far from the truth.

First, I like it when youre happy so we try to stay happy :) The F-14 is more demanding than other A/C's(Heatblurs AJS37 Viggen does about the same). Dunno why but I see they emulate vibrations and shaking for the aircraft. More than I ever saw live. For down below I use F-18 thats better than F-14 on conserving the computer power. I can add F-14 to the mission and test that one later but as f-14 is worse than other aircrafts, the F-14 is part of the problem(at least) and its not only DCS fault. The settings posted down below. I recently got the Pimax 8KX and I havent really tuned the computer for this yet, so I am in between, with a few settings lowered to keep it flowing with the Pimax. My normal settings, used with the 2080ti OC was SS150% and most often no MSAA. Otherwise I think its like the pictures below. When I started DCS MSAA wasn't enabled. I normally did use 150% in SteamVR and with supersampling is kind of does the same thing for the jagged lines. I mainly dont put the sliders all the way to the right if I cannot notice a difference or a very small diff that loads the computer much. So, my regular setting was SS150 and as below but not MSAA at all. With the setting without MSAA I kept a rock steady 90fps on the SC. Except when starring at the command bridge. I had not notice it until you guys wrote about this. If I enable MSAA 2X I still have 90fps, but it start hunting from time to time between 89 and 91 so I gues its close to not keeping 90fps. Still rock solid smooth though. I tried to shoot a film with geforce experience but then I actually did get chaning between 90 and 45fps, so it took my computer on the other side of the limit. And, how ever I do, if I shoot a movie at 60fps it wont come out smooth despite having the game flowing nice at 90fps, so I think some will get stuck seeing that. I have no mods at all. resolution according to the steamvr setting below. Other settings as below. The CPU is clocked to 5.2Ghz on all cores and the 3090 is set to the default overclock mode in the asus GPU-tweak. I have tweaked/overclocked it further, and this might give med a few extra percent but as others might not have I did run it int the asus GPU tweak Overclock mode.

What I ment was that the term allocated memory is used for shared RAM (shared between CPU and GPU). I'd say that the number you see is close to the actually used VRAM. If you look into the task manager and use the performance tab the GPU memory you see is the one reported from the VidMm that is the windows process managing the VRAM. At least that was what the windows developers told us on their blog when the update that got us the GPU/VRAM load/usage to the task manager. So the idea that something allocates a lot of video memory but only a small part is used is probably not true. I often get 17-19Gb usage reported in fpsvr with the pimax 8KX, and slightly more in task manager. Probasbly because windows use a small amount for the windows screen seen.

Are you sure theres really any allocating of memory on a for VRAM dedicated memory ? Id say it uses 12Gb.

I think some say StramVR might be less demanding than Pimax. Also, I will initially try to use low supersampling to keep the load low.

I have about the same system, 9900KS@5.2 32Gb@3200, asus strix 3090 OC. I never sen the CPU bottleneck. In fact, when using my 8KX I get quite low frame times on the CPU and the GPU really get to work hard. I dont remember the exakt values but I think I often have GPU times three times higher than the CPU. Say, GPU 25 and CPU 6 to 8 ms, I would say that you should not use PD in dcs. It use more cpu power and use pitool ss or steamvr SS instead, as these use tge GPU for the same thing. I havent got my base stations yet so I dont use the 8KX that much yet. I see in STEAMVR that the default settings in resolution per eye is very high with the 8KX. How do you know that it is the CPU that is the bottleneck? Do you use fpsvr ? For the 8KX, I dont se much mura. Doesnt notice it. No other quality problems found yet for me.