Jump to content

SYN_Vander BENCHMARK v6 to measure IL-2 performance in monitor & VR


chiliwili69
 Share

Recommended Posts

2 hours ago, chiliwili69 said:

Voxman went from 2200 to 4400MHz in RAM and got +8fps in the CPU test. 3200 is not a bad frequency for CL16.

I'm still running tests as the 4400MHz has not been stable.  I've contacted Patriot memory tech support to see if I can get to the rated amount without freeze ups and other issues.  The best frequency and stability has been 3900Mhz.  

 

I've overclocked my CPU as well, but I'm still trying to get my head wrapped around all the settings in BIOS.  Throttling of CPU and all those little factory Turbo settings will affect stability and I need to understand this stuff.  As I mentioned 9 years ago was my last build and was a pretty easy system to OC as there weren't that many settings in BIOS to manipulate. 

Nevertheless, all this is taking time, and of course I want to play the game.  I will say that currently and with the 3900Mhz settings, and a slight OC on the CPU I'm seeing FPS well above 120 in 1440 resolution, but at custom graphic settings in game.   

 

Link to comment
Share on other sites

3 hours ago, chiliwili69 said:

Many thanks for these test. It is the first 3090 tested here and the first G2 as well. You have a pretty good system.

 

The 155.6 fps in 4K test it is amazing!! This is really a good card! (thanks Mrs Charlo!)

 

The 74 in VR is also good, although I believe that the limitation is not the GPU.


Your past posts helped guide me in this build, thank you! I’ll be curious to see if/when Steam improves their handling of the G2 resolution in SteamVR. In actual MP gameplay I am still experimenting with the optimal resolution to maintain 90 fps on Finnish and Combat Box. I would also like to lower how shimmery the wires of other planes are in Flying Circus quick missions.

Link to comment
Share on other sites

4 hours ago, nickj123 said:

I'm not too concerned as it could be something with the AMD card that is causing the slowdown.

 

Do you have another GPU? (your previous one, just to check it doesn´t affect the CPU test).

 

You can also can do two test:

- Try other XMP profiles for your RAM, even lower frequency

- Just reduce the resolution in the CPU test, for example 1024x768, so not impacted by GPU slowdown.

Link to comment
Share on other sites

On 11/24/2020 at 4:09 AM, SCG_Fenris_Wolf said:

 Motherboard: MSI X570 Tomahawk WIFI
 CPU:             AMD Ryzen 9 5900X
 CPU Freq:        4950 MHz
 L3 cache:        2x 32 MBytes
 Cores:           12
 Threads:         24
 RAM type:        DDR4
 RAM size:        2 x 16 GB (Dual Rank)
 Uncore Freq:     1900 MHz FCLK==MCLK
 RAM Freq:        3800 MHz
 RAM Latency:     16-16-16-16-32 
 GPU:             MSI RTX 3080 Trio X Gaming (TDP 102%, +100MHz core, + 250MHz memory)

 

 

CPU Test (PBO disabled, stock speed)

2020-11-23 18:47:27 - Il-2
Frames: 7668 - Time: 60000ms - Avg: 127.800 - Min: 111 - Max: 175

 

GPU Test

<place holder> //DSR stretches image beyond screen. Will test tonight in my living room, got a 4K TV there

 

VR Test

<place holder>

 

hi there, Wolf,

 

i'm very much interested in your build.  Did you finally happen to run the VR benchmark test ?

 

Link to comment
Share on other sites

2 hours ago, Voxman said:

Nevertheless, all this is taking time, and of course I want to play the game.  I will say that currently and with the 3900Mhz settings, and a slight OC on the CPU I'm seeing FPS well above 120 in 1440 resolution, but at custom graphic settings in game.

 

That´s why it is strange that you are below 100 with just 1080p resolution.

You and nickj123 have both good intel CPU and should be above 100 in CPU test. And you are the only ones with Intel CPU and AMD GPU. That´s perhaps the reason?

In any case, playing IL-2 is more fun than tweaking. Tweaking is something one want to do once and forget.

Link to comment
Share on other sites

Motherboard: Asus Crosshiar VIII Hero X570 (Bios 2702)

CPU: AMD Ryzen R9 5950X

CPU Frequency: 4.775GHz

Level 3 Cache: 64MB

Cores: 16

Threads: 32

Ram Type: DDR4

Ram Size: 2 x 8GB

Uncore Frequency: 1866.8

Ram Frequency: 2 x 3733MHz

Ram Latency: 16 16 16 32 CR1

GPU: AMD Vega 64

 

cachemem-3733-pbo13-32.thumb.png.e50c3b296b5d5d679847f9ec98e1cd97.png

 

20201203_174729.thumb.jpg.b32733ba8b5e14ff7007c4171babbe67.jpg

 

CPU:
2020-12-02 22:10:18 - Il-2
Frames: 6327 - Time: 60000ms - Avg: 105.450 - Min: 92 - Max: 154

 

GPU:

2020-12-03 16:48:48 - Il-2
Frames: 1762 - Time: 60000ms - Avg: 29.367 - Min: 25 - Max: 35

 

Out of curiosity I changed from MSAA to FXAA which forced settings for X8 to X4 and got this:

2020-12-03 16:52:08 - Il-2
Frames: 7131 - Time: 60000ms - Avg: 118.850 - Min: 90 - Max: 140 - Still 4K :)

 

I have a Rift CV1 - what settings should I set that at to match the resolution the benchmark is chasing?

 

 

Il-2 2020-12-02 22-10-18-10 fps.zip Il-2 2020-12-03 16-48-48-18 fps.zip

 

Edited by robbiec
Link to comment
Share on other sites

1 hour ago, robbiec said:

Out of curiosity I changed from MSAA to FXAA which forced settings for X8 to X4 and got this:

2020-12-03 16:52:08 - Il-2
Frames: 7131 - Time: 60000ms - Avg: 118.850 - Min: 90 - Max: 140 - Still 4K :)

 

I have a Rift CV1 - what settings should I set that at to match the resolution the benchmark is chasing?

 

For the Rift CV1 to have per eye 1948x2319=9million pixels you can either put 210% SteamVR_SS or put PixelDensity 1.45 in OTT/ODT (1.45 is just Squareroot of 210/100)

 

When you get 105 fps in the CPU test most of us thought that it was due to the Vega64 card. In theory a Vega64 is on pair with an 1080 card. So this should not explain your low value in the CPU test.

 

When you made the GPU 4K test with just just FXAA (all other settings as the GPU test) then you achieved 119 in 4K, so it means that your card can manage well 4K (8.3million pixels) when MSAA is not used.

So, your Vega64 can not be the bottleneck when you run the CPU test since it doesn´t use AA at all and 1080p is just only 2 million pixels. You can try to put a very low resolution (1024x768) and most likely you will get your 105 mark in the CPU test.

 

You should deliver a fps on pair with Fenris mark for the CPU test. Only the dual rank thing can not explain this 22fps difference.

Link to comment
Share on other sites

JimmySolarium_VR

Motherboard: MSI X570-Gaming PLUS
 CPU:                 Ryzen 5 3600X
 CPU Freq:        4.07 Ghz
 L3 cache:        2x16 MB
 Cores:               6
 Threads:           12
 RAM type:        DDR4
 RAM size:        16 GB
 NB Freq:          1600 MHz
 RAM Freq:        3200 MHz
 RAM Latency: 14-14-14-34
 GPU:                 RTX 2070 S


CPU Test
2020-12-03 21:20:23 - Il-2
Frames: 4810 - Time: 60000ms - Avg: 80.167 - Min: 68 - Max: 112

 

VR Test Oculus CV1 OTT: 1.4

2020-12-03 21:31:12 - Il-2
Frames: 2702 - Time: 60000ms - Avg: 45.033 - Min: 42 - Max: 53

 

Some results of my actual machine. Awaiting a Reverb G2 but I fear i have to upgrade a lot ....

Cheers,

Jimmy

Edited by JimmySolarium
redone
Link to comment
Share on other sites

1 hour ago, chiliwili69 said:

You can try to put a very low resolution (1024x768)

 

I have been running the CPU test as the instructions (using 1920x1080 res) and then also tried to run it with a lower res with same proportion to have the same FoV (I used 1280x720 IL-2 resolution). My monitor is 4K, so I don´t use full screen mode.

As you can see the results are very similar altough not identical:

 

CPU Test at 1920x1080
Frames: 4818 - Time: 60000ms - Avg: 80.300 - Min: 70 - Max: 109
Frames: 4786 - Time: 60000ms - Avg: 79.767 - Min: 69 - Max: 109
Frames: 4783 - Time: 60000ms - Avg: 79.717 - Min: 71 - Max: 108
Frames: 4870 - Time: 60000ms - Avg: 81.167 - Min: 72 - Max: 110

 

CPU test at 1280x720
Frames: 4988 - Time: 60000ms - Avg: 83.133 - Min: 73 - Max: 105
Frames: 4829 - Time: 60000ms - Avg: 80.483 - Min: 67 - Max: 103
Frames: 4930 - Time: 60000ms - Avg: 82.167 - Min: 72 - Max: 109

 

With this, I would want to demostrate that people who think that the GPU could be limiting the CPU test, they can run the test at lower resolution (1280x720 is the lower for the same proportion) just to find that something else could be happening in the CPU or in the RAM or in the bus, but not related to the power of the GPU.

Link to comment
Share on other sites

2 hours ago, JimmySolarium said:

I fear i have to upgrade a lot ....

Thanks for your results. The good thing is that you can keep the Mobo and the RAM if you go to the new Zen3.

Link to comment
Share on other sites

SCG_Fenris_Wolf
6 hours ago, 69th_Didney_World said:

 

hi there, Wolf,

 

i'm very much interested in your build.  Did you finally happen to run the VR benchmark test ?

 

Hi, I've put it on my plan for tomorrow

  • Like 1
  • Upvote 1
Link to comment
Share on other sites

3 hours ago, chiliwili69 said:

When you get 105 fps in the CPU test most of us thought that it was due to the Vega64 card. In theory a Vega64 is on pair with an 1080 card. So this should not explain your low value in the CPU test.

 

You should deliver a fps on pair with Fenris mark for the CPU test. Only the dual rank thing can not explain this 22fps difference.

 

Well dual rank is one variable, other is that he has managed to have 8C at a solid 4950 (which is a really, really good overclocking result - actually a spectacular result without LN) The nearest I can get to that is 4800 and his Infinity Fabric is 1900 verses my 1866 - all add up. Extra memory banks will probably take the memory Read & Write to 58GB/s but I'm expecting the COPY to jump from 49 to mid 50s.

As you can see by the AIDA memory scores, my memory is in reasonable tune now and my Cinebench scores are as expected. So short of a fresh Win 10 install, there is not much more I can do. 

I think I've managed to get hold of a RX6800, will be confirmed tomorrow. 

 

Thanks for the CV1 settings, appreciated. S! 

Link to comment
Share on other sites

SCG_Fenris_Wolf

Yes, small nitpicking clarification, I didn't overclock the CPU, the 5900X boosts to 4950MHz out of the box, it performs best at stock settings (better than even with PBO).

 

The RAM quality is indeed very important for IL-2. I have ~52ns latency. Dual Rank is very important especially once you move across the map instead of staying in the same grid, to avoid frametime spikings and performance issues.

 

What I read from your post is that you run the infinity fabric at 1866MHz and the memory at 4800MHz. If that is the case, I can say with certainty that this is a major reason for the performance difference: Your IF (FCLK) must be run the same frequency as your memory controller (MCLK).

If your Fclk is at 1866MHz, your RAM must have a speed of 3733MHz. If you set it to higher frequency, you system will be slower. What happens is that the timings are off, and the system will "stumble over its own feet all the time".

 

Note I used "frequency", not speed. Speed is the result of "having the same frequency, and then having tightened the timings" (CL16 -> CL15, etc).

 

Also, if you have switched motherboards, you will need to run a format C:. Also make sure to freshly install AMD's chipset drivers (from AMD website, not the motherboard manufacturers'!).

 

TLDR:

1. FCLK==MCLK: Infinity Fabric at 1866MHz requires RAM at 3733MHz, or system runs slower (with frametime spikes!)

2. Dual Rank RAM is important, or system runs slower (with frametime spikes especially while travelling!)

3. If you have switched the motherboard, format C: and reinstall Windows clean. Get AMD's chipset drivers too.

Edited by SCG_Fenris_Wolf
  • Thanks 1
  • Upvote 1
Link to comment
Share on other sites

20 minutes ago, SCG_Fenris_Wolf said:

and the memory at 4800MHz.

He refers 4800 for the CPU clock, no memory.

 

In his result post he say that RAM runs at 2x3733. I understand it is just 3733.

Link to comment
Share on other sites

SCG_Fenris_Wolf

Ah alright.

 

Well then it might be useful for other newcomers to the ZEN platform. Many are to come I suppose.

Link to comment
Share on other sites

28 minutes ago, SCG_Fenris_Wolf said:

3. If you have switched the motherboard, format 😄 and reinstall Windows clean.

 

BTW, I will go through a MoBo switch when my new 5600X will arrive (hopefully in a week) and I have already acquired a Western Digital SSD M.2 Sata3 2TB unit as my future and only storage device. Currently I have on 240GB SSD (with Windows and IL-2) and a 1TB HD. These two devices will go away.

Maybe a silly question, Can I keep my Windows license? and what is the best procedure for a fresh install?

Link to comment
Share on other sites

SCG_Fenris_Wolf

Yes, you can, you need to tie it to your Microsoft Account if you have not yet. Log-in to Microsoft's Website and check your products, you'll also find your license there.

 

Then you can keep the license. Just press "skip product key" during installation. It's important you select the precise type of Win10 (Pro, 64, or Home, etc) that your license is set to.

 

After start-up, you can login to your Microsoft Account. There's no way around the account afaik =]

Edited by SCG_Fenris_Wolf
Link to comment
Share on other sites

2 hours ago, SCG_Fenris_Wolf said:

Yes, small nitpicking clarification, I didn't overclock the CPU, the 5900X boosts to 4950MHz out of the box, it performs best at stock settings (better than even with PBO).

 

The RAM quality is indeed very important for IL-2. I have ~52ns latency. Dual Rank is very important especially once you move across the map instead of staying in the same grid, to avoid frametime spikings and performance issues.

 

What I read from your post is that you run the infinity fabric at 1866MHz and the memory at 4800MHz. If that is the case, I can say with certainty that this is a major reason for the performance difference: Your IF (FCLK) must be run the same frequency as your memory controller (MCLK).

If your Fclk is at 1866MHz, your RAM must have a speed of 3733MHz. If you set it to higher frequency, you system will be slower. What happens is that the timings are off, and the system will "stumble over its own feet all the time".

 

Note I used "frequency", not speed. Speed is the result of "having the same frequency, and then having tightened the timings" (CL16 -> CL15, etc).

 

Also, if you have switched motherboards, you will need to run a format C:. Also make sure to freshly install AMD's chipset drivers (from AMD website, not the motherboard manufacturers'!).

 

TLDR:

1. FCLK==MCLK: Infinity Fabric at 1866MHz requires RAM at 3733MHz, or system runs slower (with frametime spikes!)

2. Dual Rank RAM is important, or system runs slower (with frametime spikes especially while travelling!)

3. If you have switched the motherboard, format 😄 and reinstall Windows clean. Get AMD's chipset drivers too.

 

Ram = 3733 -16-16-16-32, FBLK = 1866 - I'll perform a fresh Win10 install and put CPU at stock and see what happens. Got an order confirmation for RX6800 :)

Link to comment
Share on other sites

5 hours ago, chiliwili69 said:

 

BTW, I will go through a MoBo switch when my new 5600X will arrive (hopefully in a week) and I have already acquired a Western Digital SSD M.2 Sata3 2TB unit as my future and only storage device. Currently I have on 240GB SSD (with Windows and IL-2) and a 1TB HD. These two devices will go away.

Maybe a silly question, Can I keep my Windows license? and what is the best procedure for a fresh install?

 

adding to what @SCG_Fenris_Wolf already said you can check the below article out if you're using a license with a product key, there are couple of commands to remove the key from the local installation. However it's really only applicable if you have both the OLD machine and the NEW machine running at the same time. If you're not, i.e. If you're simply upgrading your machine to a diff mobo/cpu/memory then the commands mentioned in the article aren't needed. The article is still worth looking at as it shows you the actual local windows commands for license management.

 

https://www.windowscentral.com/how-transfer-windows-10-license-new-computer-or-hard-drive

 

 

Edited by 69th_Didney_World
  • Thanks 1
Link to comment
Share on other sites

38 minutes ago, 69th_Didney_World said:

dding to what @SCG_Fenris_Wolf already said

 

Thanks. I checked and I have a Digital license of windows associated to my microsoft account (I didn´t know it!).

 

I will not use the old Mobo/RAM/CPU, will sell it through ebay.

  • Upvote 1
Link to comment
Share on other sites

thermoregulator

Ryzen 9 5900X, stock settings, FCLK 1900 Mhz

Arctic liquid freezer II 360, push-pull config

MSI PRESTIGE X570 CREATION

RAM G-SKILL F4-3200C14D-32GTZN, 32 GB dual rank, 3800 Mhz, 1T, CL 16-16-16-32, tRC 48, tRFC 304

EVGA RTX 3090 FTW3 Ultra Gaming

HP Reverb G1, 100% SteamVR SS (2204 x 2164)

 

VR Benchmark, first run:

 

2020-12-04 21:08:44 - Il-2
Frames: 5041 - Time: 60000ms - Avg: 84.017 - Min: 61 - Max: 92

 

 

VR Benchmark, second run (tweaked RAM subtimings a little)

 

2020-12-04 22:10:20 - Il-2
Frames: 5173 - Time: 60000ms - Avg: 86.217 - Min: 69 - Max: 91

 

VR Benchmark, third run:

 

2020-12-04 22:45:08 - Il-2
Frames: 5294 - Time: 60000ms - Avg: 88.233 - Min: 78 - Max: 91

(I found out that during first two runs, I had lower screen resolution, just HD. As I raised the screen resolution to FHD as suggested in OP, strangely I get better  FPS results in the third run)

Edited by thermoregulator
Link to comment
Share on other sites

2 minutes ago, thermoregulator said:

RAM G-SKILL F4-3200C14D-32GTZN, 32 GB dual rank, 3800 Mhz, 1T, CL 16-16-16-32, tRC 50, tRFC 304

 

you oc'ed the memory to 3800mhz to get to 16 CAS latency?

 

Did you try running this benchmark using the standard 3200mhz 14CAS timings?

Link to comment
Share on other sites

thermoregulator
17 minutes ago, 69th_Didney_World said:

 

you oc'ed the memory to 3800mhz to get to 16 CAS latency?

 

Did you try running this benchmark using the standard 3200mhz 14CAS timings?

I just used DRAM calculator suggested 3800 Mhz timings for my memory, I didn't have time to play with it yet. 

  • Upvote 1
Link to comment
Share on other sites

SCG_Fenris_Wolf

VR test results are out

 

 

P.S. I found out, from Joanna Popper on Reddit (HP VR Lead), that the Reverb G2 will require more than 9.3Mio pixels to run native resolution. SteamVR shows the framebuffer, which includes the distortion profile. I estimate that we'd need at least 1.1x for the resolution to be proper. That's above 10Mio, the 6800XT reportedly suddenly stutter themselves to death above 9Mio, while the 3080 really sweat a lot. This might need a 3090 to play at around 90Hz while using supersampling or msaa. I wonder how the 6900XT performs, it's on my list as well! 😇 All my hopes rest on the 6900XT. Don't get me wrong, the 3080 holds very well too, and I think the 3090 is just way too far out of line regarding price. 

Edited by SCG_Fenris_Wolf
  • Thanks 2
  • Upvote 1
Link to comment
Share on other sites

41 minutes ago, SCG_Fenris_Wolf said:

The Reverb G2 will require more than 9.3Mio pixels to run native resolution. SteamVR shows the framebuffer, which includes the distortion profile. I estimate that we'd need at least 1.1x for the resolution to be proper. That's above 10Mio, the 6800XT reportedly suddenly stutter themselves to death above 9Mio, while the 3080 really sweat a lot. This might need a 3090 to play at around 90Hz while using supersampling or msaa. I wonder how the 6900XT performs, it's on my list as well! 😇 All my hopes rest on the 6900XT. Don't get me wrong, the 3080 holds very well too, and I think the 3090 is just way too far out of line regarding price. 

 

Hmm, a "sudden stutter" sounds like running out of VRAM, but the 6800XT has 16GB so it shouldn't be that. There are rumors of a 3080ti 20GB in January, as well as a 3070ti 10GB coming. If a 3080ti appears I will use EVGA step-up to upgrade to it.

 

Do we know the proper 100% SteamVR resolution for G2 yet? I have one on back order, no idea when it will arrive, but I would like to set up the rest of my system with correct settings, see if I can hit 90hz, etc etc.

Link to comment
Share on other sites

SCG_Fenris_Wolf
50 minutes ago, Alonzo said:

 

Hmm, a "sudden stutter" sounds like running out of VRAM, but the 6800XT has 16GB so it shouldn't be that. There are rumors of a 3080ti 20GB in January, as well as a 3070ti 10GB coming. If a 3080ti appears I will use EVGA step-up to upgrade to it.

 

Do we know the proper 100% SteamVR resolution for G2 yet? I have one on back order, no idea when it will arrive, but I would like to set up the rest of my system with correct settings, see if I can hit 90hz, etc etc.

 

Here are some very interesting GPU Benchmarks in VR https://babeltechreviews.com/category/virtual-reality/. I am thrilled how the 6900XT may perform in VR especially. We can't use synthetic benchmarks or 2D benchmarks of the Youtuber-horde to get a precise enough idea on how it'll perform for us.

Edited by SCG_Fenris_Wolf
Link to comment
Share on other sites

54 minutes ago, Alonzo said:

 

Do we know the proper 100% SteamVR resolution for G2 yet? 

 

I have been wondering this as well.  It seems like there is some conflicting information on what is a best practice.

Link to comment
Share on other sites

SCG_Fenris_Wolf

It's been said that the default SteamVR resolution (100%) wasn't the proper one yet, and may change with consecutive updates. We ought to ask Miss Popper for the proper target resolution which we'd need to get to the panels natively including of what the distortion profile needs. Maybe she'll reply if it comes up regularly. She's active on https://www.reddit.com/r/HPReverb/

 

My G2 is shipping right now, so we'll know soon enough regarding what looks sufficient without pixel crawl. 

Edited by SCG_Fenris_Wolf
Link to comment
Share on other sites

Did some dirty quick OC....

 

Motherboard: msi meg z490 unifz
 CPU:                i7 10700k
 CPU Freq:        5.0Ghz
 L3 cache:        16 MB
 Cores:               8
 Threads:           16
 RAM type:        DDR4
 RAM size:        32 GB
 RAM Freq:       3200 MHz 
 RAM Latency: 16
 GPU:                 rtx 2070super

 

CPU TEST

2020-12-05 06:23 - Il-2
Frames: 5805- Time: 60000ms - Avg: 96.750 - Min: 85 - Max: 128

 

Edit: in a week or so once my 3080 arrives i'll do benchmark on VR- rift s......G2 will have to wait January.

P.S. darn that amd 5x00x series are fast, and damn Reverb G2 with their usb mobo power issues 😬

Edited by =VARP=Ribbon
Link to comment
Share on other sites

On 12/4/2020 at 9:27 PM, thermoregulator said:

2020-12-04 22:45:08 - Il-2
Frames: 5294 - Time: 60000ms - Avg: 88.233 - Min: 78 - Max: 91

 

Wow! That´s a very good mark. I didn´t thought that we were going to be so close to 90fps in this test.

Could you also please run the CPU test? and GPU test is 4K available?

 

I have realized that the instructions were not very clear for the SS% to be applied, since 100% in Index is 9 million, but 100% in G1 is 9.54 million. So I have modified the instructions indicating the % that should be used for every VR device, taking as the base the 9.54million.

The table introduced in  the procedure is now:

 

Reverb G1  SS=100%  2205x2165=9.54 Mpixels  @nickj123

Reverb G2  SS=54%   2205x2165=9.54 Mpixels (until HP correct that)  @Charlo-VR
Index           SS=106%  2076x2306=9.57 Mpixels   @R3animate 
RiftS            PD=1.36  2235x2409=10.77Mpixels at 80Hz (184% SS)

Pimax8KX  SS=78%   2219x2173=9.64 Mpixels (90HZ, with Normal FOV, Render Quality=1) 
Pimax8K     SS=38%   2346x2066=9.69 Mpixels (90HZ, with Normal FOV, Render Quality=1)
Pimax5K+  SS=38%   2375x2029=9.63 Mpixels (90HZ, with Normal FOV, Render Quality=1)
VivePro      SS=106%  2076x2306=9.57 Mpixels 
Odyssey     SS=188%  1958x2436=9.54 Mpixels
old WMR    SS=188%  2184x2184=9.54 Mpixels
Rift CV1     PD=1.49  1882x2240=9.55 Mpixles (222% SS) @JimmySolarium   @robbiec
Vive            SS=188%  2073x2303=9.55 Mpixels

Edited by chiliwili69
  • Thanks 1
Link to comment
Share on other sites

On 11/24/2020 at 10:09 AM, SCG_Fenris_Wolf said:

2020-12-04 22:14:55 - Il-2
Frames: 5346 - Time: 60000ms - Avg: 89.100 - Min: 83 - Max: 92

 

WOW!!. This togheter with the another result of thermoregulator is quite nice. So, it seems that either 3080 or 3090 with new ryzens reach the nirvana!!

One question? which VR device did you use? 8KX? at 90Hz?

 

You had some dips below 90, do you think it was CPU or GPU contrained?

 

I am thinking that maybe the settings used in the VR test are a bit low. Do you think they will need to be raised a bit so top rigs are still a bit more below?

Link to comment
Share on other sites

thermoregulator
1 hour ago, chiliwili69 said:

 

Wow! That´s a very good mark. I didn´t thought that we were going to be so close to 90fps in this test.

Could you also please run the CPU test? and GPU test is 4K available?

 

I have realized that the instructions were not very clear for the SS% to be applied, since 100% in Index is 9 million, but 100% in G1 is 9.54 million. So I have modified the instructions indicating the % that should be used for every VR device, taking as the base the 9.54million.

The table introduced in  the procedure is now:

 

Reverb G1  SS=100%  2205x2165=9.54 Mpixels  @nickj123

Reverb G2  SS=54%   2205x2165=9.54 Mpixels (until HP correct that)  @Charlo-VR
Index           SS=106%  2076x2306=9.57 Mpixels   @R3animate 
RiftS            PD=1.36  2235x2409=10.77Mpixels at 80Hz (184% SS)

Pimax8KX  SS=78%   2219x2173=9.64 Mpixels (90HZ, with Normal FOV, Render Quality=1) 
Pimax8K     SS=38%   2346x2066=9.69 Mpixels (90HZ, with Normal FOV, Render Quality=1)
Pimax5K+  SS=38%   2375x2029=9.63 Mpixels (90HZ, with Normal FOV, Render Quality=1)
VivePro      SS=106%  2076x2306=9.57 Mpixels 
Odyssey     SS=188%  1958x2436=9.54 Mpixels
old WMR    SS=188%  2184x2184=9.54 Mpixels
Rift CV1     PD=1.49  1882x2240=9.55 Mpixles (222% SS) @JimmySolarium   @robbiec
Vive            SS=188%  2073x2303=9.55 Mpixels

I tried to run both CPU and GPU test (i have 4K 60 Hz freesync screen). But even though I have disabled adaptive sync and v sync in NvCP, I was still locked on 60 Hz in fraps. I am gonna try to solve this later. 

Link to comment
Share on other sites

7 hours ago, =VARP=Ribbon said:

Frames: 5805- Time: 60000ms - Avg: 96.750 - Min: 85 - Max: 128

 

Thank you for your test. You also can run the test in VR with the Rift-S and the 2070S. So you will see the improvement with the new 3080.

BTW, according to WalteScott he didn´t see performance increase beyond 5.0GHz. But he was running the RAM at 4400MhZ with CL16.

  • Upvote 1
Link to comment
Share on other sites

thermoregulator
22 minutes ago, chiliwili69 said:

 

WOW!!. This togheter with the another result of thermoregulator is quite nice. So, it seems that either 3080 or 3090 with new ryzens reach the nirvana!!

One question? which VR device did you use? 8KX? at 90Hz?

 

You had some dips below 90, do you think it was CPU or GPU contrained?

 

I am thinking that maybe the settings used in the VR test are a bit low. Do you think they will need to be raised a bit so top rigs are still a bit more below?

I will try to run it on my usual settings tomorrow.

Link to comment
Share on other sites


 Motherboard: EVGA Z390 Dark
 CPU:                 i9 9900k
 CPU Freq:        5.2 GHz all cores HT off
 L3 cache:        
 Cores:               8 (number of active physical Cores)
 Threads:          8
 RAM type:        DDR4
 RAM size:        32Gb (2x16)
 NB Freq:         
 RAM Freq:       3200 MHz
 RAM timings:  CL 14
 GPU:                 EVGA 3090 FTW3 Ultra

 CPU Cooler: Corsair H150i Pro AIO
 RAM Model: G Skill Trident Z 3200 MHz CL14.

 

1080p Monitor Test :

Frames: 6397 - Time: 60000ms - Avg: 106.617 - Min: 90 - Max: 141

 

VR Test - HP Reverb G2:

Frames: 4303 - Time: 60000ms - Avg: 71.717 - Min: 55 - Max: 91

 

Both tests run with specified GFX settings for each one.

 

Edit: I realized after the fact I had SSAO enabled and in the OP listing it was to not be enabled.

Not sure if would make any difference, may re-run the test later and will update this post if it does.

Edit: 2 -  I went ahead and disabled SSAO and reran the test, made virtually no difference results the same.

 

 


 

Edited by dburne
Link to comment
Share on other sites

5 hours ago, chiliwili69 said:

I have realized that the instructions were not very clear for the SS% to be applied, since 100% in Index is 9 million, but 100% in G1 is 9.54 million. So I have modified the instructions indicating the % that should be used for every VR device, taking as the base the 9.54million.

The table introduced in  the procedure is now:

 

Reverb G1  SS=100%  2205x2165=9.54 Mpixels  @nickj123

Reverb G2  SS=54%   2205x2165=9.54 Mpixels (until HP correct that)  @Charlo-VR


I’m happy to run my VR test again but am still unclear about the SteamVR settings for SS. As a former Quest user and non-Steam user, I’ve only been using SteamVR for a few weeks since I received my G2. Last night I notice that SteamVR’s resolution settings are more complicated than I had thought.

 

There are two places resolution is controlled in SteamVR - on the General tab (which is repeated on the Video tab), and on the per-application tab (available from the Video tab). The per-application tab’s tooltip explains the setting on that tab is a multiplier of the resolution on the General tab.
 

For my next VR test do I set both the resolution on the General tab and on per-application tab to 54% (which is  2205x2165)?

 

How does the math for those two tabs’ settings work together to get the best resolution for the G2? My Google searching didn’t provide me a clear explanation of that as I try to dial in my SteamVR settings to reduce the shimmering and keep 90 fps. 

Link to comment
Share on other sites

2 hours ago, dburne said:

VR Test - HP Reverb G2:

Frames: 4303 - Time: 60000ms - Avg: 71.717 - Min: 55 - Max: 91

 

Many thanks Don for running the test with the G2 and the 3090. You have a good beast as well. I believe that the 3090 is not fully loaded at any moment during the test.

Yes, the SSAO thing doesn´t have an impact. At least for CPU.

2 hours ago, Charlo-VR said:

For my next VR test do I set both the resolution on the General tab and on per-application tab to 54% (which is  2205x2165)?

 

That´s a good question. I have never used on per-application tab in SteamVR (in fact IL-2 doesn´t appear there as it is not IL-2 from Steam).

I only use the General or Video tab which are the same.

 

I believe that you should leave on per application setting at 100%, so the value specified in the General tab is not affected when running IL-2. I mean 54%*100% is 54%.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...