(Il-2 Great Battles) SYN_Vander BENCHMARK v6 to measure IL-2 performance in monitor & VR by chilliwilli69

[LuftManu]

This is a benchmark procedure targeted to measure independently the fps performance of different CPUs/RAMs and different GPUs when running IL-2 in a monitor (and VR)
The benchmark is based in a mission built by SYN_Vander. It is a 60 seconds low altitude flight over dense and beautiful train station of Velikie Luki summer map. An external camera view behind an IL-2 and some more planes around. It is a quite heavy scenario.

 

The purposes of having a benchmark for IL-2 are:
- To know if your hardware&software is performing as expected (comparing to peers)
- To know the best CPUs/RAM and GPUs for IL-2.
- To know the performance cost of every graphic setting and the trade-offs (visual gain vs performance)

 

There are two test for monitor and three for VR:
1. CPU test in monitor: Run in 1920x1080 (FullHD) resolution. Here the bottleneck should be the CPU. (GPU load should be always well below 100%)
2. GPU test in 4K monitor: Run in 3840x2160 (4K) resolution. Here the bottleneck should be the GPU.
3. CPU&GPU&VR-device test in VR: Settings for 9.54 Million pixels at 90Hz. There are many factor affecting VR performance (device, software ecosystem, FOVs, panel refresh rate, etc)

4. CPU&GPU&VR-device test in VR: Settings for 19.5 Million pixels at 90Hz.

5. CPU&GPU&VR-device test in VR: Settings for 19.5 Million pixels at 90Hz with Extreme clouds and MSAAx8.

 
You can follow the instructions below to run the two tests in your PC with monitor. And if you have VR, you can also run the third&fourth test.

I will compile all collected info and will update it in the below spreadsheet, in SYN_Vander tab:

https://docs.google.com/spreadsheets/d/1gJmnz_nVxI6_dG_UYNCCpZVK2-f8NBy-y1gia77Hu_k

 

If your test results are much lower than other tests similar to your PC you can review these items:

https://forum.il2sturmovik.com/topic/34107-items-to-review-if-you-have-low-performance-in-il-2-benchmark/

=================================================================================================

Instructions to run the SYN_Vander benchmark:
Note: Before running this test you should first spend the required time to Overclock your CPU to the maximum (and stable) capacity of your PC. There is multiple info on the web for Overclock.
Be sure to run all your cores at the same manually fixed CPU frequency and set he AVX Offset=0 . So all the tests performed are run at the same CPU frequency with no CPU throttling.

 

1. GET YOUR PC INFO
 
Install and execute the free tool CPU-Z from: https://www.cpuid.com/softwares/cpu-z.html
Collect the following info from CPU-Z:

 

Motherboard: Gigabyte B550 Aorus Pro
 CPU:                 Ryzen 5 5600X
 CPU Freq:        4.65 Ghz
 Cores:               6 (number of active physical Cores)
 Threads:           12 (HyperThreading is better off. If Cores=Threads then HT is Off)
 RAM size:        32Gb (2x16GB)
 RAM Freq:       3600 MHz (this is 2xDRAM freq with Dual channel)
 NB Freq:          1800 MHz (or Uncore Frequency or UCLK)
 RAM timings:  16-16-16-36
 GPU:                 3080

2. INSTALL measuring tool (NEW Tools are being considered):

Install Fraps (free fps measuring software) from: http://www.fraps.com/

Run Fraps. 

Under the "general" tab, uncheck "fraps window always on top". 

Under the "99 fps" tab, set benchmarking hotkey to "P". 

Also check and set the "stop benchmark after __ seconds" to 60 seconds

3. INSTALL SYN_Vander benchmark mission:
 

Get the mission benchmark zip file. Download it from here: https://drive.google.com/file/d/1zKyuQhVRxRysjXhOnlsniLa-V73FMCd4/view?usp=sharing

 

Unzip it and drop the files into your "IL-2 Sturmovik Battle of Stalingrad\data\Missions" folder as shown below:

4. GPU Panel settings and GENERAL IL-2 Settings.

 

Turn off any IL-2 Mod or extra shader to run the test.

If you use a 4K or 2K monitor, put the Windows display scale at just 100% (Display Settings/scale).

If your GPU uses Nvidia G-Sync or AMD Free-sync, you must switch it off to don´t limit fps during the test.

In the game Settings/Flight interface  --> show HUD by default = Yes (hud enabled). Later you will press H to hidde the HUD.

Game: Map Scenery  Distance to 20Km.

Also in your NVIDIA (or AMD) control panel put the following (or closer for AMD) settings:

I use all global (I think mostly default) settings for IL-2 except Power to "Prefer Maximum performance"

5. RUN CPU TEST in 1080p

Start Fraps, if it's not already running.

Start IL-2, set IL-2 Graphics option to the below and re-boot IL-2 to make changes effective (all maxed-out except mirrors-Off, terrain roughness-off, Clouds-Low, AA-Off, SSAO-Off, HDR-Off):

 

 

 

Go to Missions, set Realism to Normal and select BenchmarkVx mission.

Press Start button. Then the map view appears and press Start button. You should see this:

Wait about 25 seconds to be sure all building are loaded in the scene.

Press "M" one or two times to hide the Map. (The IL-2 fps/FOV counter can be left on, it doesn´t affect performance). 

If Icons are shown press "I" to hide icons.

Press "P" to both: Start the Fraps measurement and unpause the mission. Then enjoy the mission run without touching anything.

 
6. RUN GPU TEST in 4K (only if you have a 4K monitor o TV)

Note: Here you need a 4K monitor or TV (using DSR factors to emulate a 4K monitor penalize the test)

Set IL-2 Graphics option to the below and re-boot IL-2 to make changes effective (all to minimum except Cloud-Extreme, MSAAx8, 4K-Textures-On) :

o to Missions, set Realism to Normal and select BenchmarkVx mission and Start button. Then the map view appears and press Start button. 

Wait about 25 seconds (load buildings)

Press "M" to hide the Map. And "I" to hide icons if shown.

Press "P" 
 

7. REPORT RESULTS.
Find the folder you installed Fraps. Inside that folder there is a "Benchmarks" folder and inside that you will find a "FRAPSLOG.TXT" file with the Frames, Avg, Min and Max fps values.

There will be also the recorded fps in *.csv files that you can open with notepad or excel if you like to plot them.

 

Report your Frames, Avg, Min, Max values for CPU and GPU test in a post with ALL your system specs from section 1 (IL-2 release, Windows version, Motherboard, CPU, CPU Freq, etc)
 

I will compile all collected info and will update it in the below spreadsheet, in SYN_Vander tab:

https://docs.google.com/spreadsheets/d/1gJmnz_nVxI6_dG_UYNCCpZVK2-f8NBy-y1gia77Hu_k

 

If your test results are much lower than other tests similar to your PC you can review these items:

https://forum.il2sturmovik.com/topic/34107-items-to-review-if-you-have-low-performance-in-il-2-benchmark/

 

8. VR TEST1 (OPTIONAL)

The VRtest1 and VRtest2 use the same IL-2 settings than the CPU test.

Set SteamVR %SS to a value to reach around 9.54 Million pixels which is the common basis (based on Reverb G1 at 100%)

If you use other non-SteamVR software like OpenComposite/OpenXR Tools/OpenXR Toolkit , adjust the SS to have around 9.5 Million pixels in total and notify that OC/OpenXR is used. Also, put Off any Upscaling/Sharpening (NIS, FSR, CAS, etc) or Foveated Rendering:

Reverb G1  SS=100%  2205x2165x2=9.54 Mpixels

Reverb G2  SS=50%    2236x2188x2=9.78 Mpixels 
Index          SS=106%  2076x2306x2=9.57 Mpixels (remember to use 90Hz mode)
RiftS           PD=1.36   2235x2409x2=10.77Mpixels at 80Hz (184% SS)

Varjo XR3  SS=28%   2167x2167x2=9.4Mpixels

Vive Pro2   Set Ultra mode (2448x2448@90Hz) and set SS to have around 2180x2180 pixels per eye.

Varjo Aero SS=56%  2350x2015=9.46 Mpixels with High Settings (default) in Varjo base software

Pico4         SS=64% 2160x2160x2=9.51Mpixels (remember to use 90Hz mode)

Quest3      SS=128% 2116x2280x2=9.64Mpixels (remember to use 90Hz mode, and 1.0 resolution in Oculus)

Crystal      SS=22% 2023x2392x2=9.67Mpixels (use 90Hz and custom resolution=1.0)

BSB1&2     SS=68% 2200x2200x2=9.68Mpixels (use 90Hz)

Pimax8KX  SS=40%   2350x1999x2=9.4 Mpixels (75HZ, with Normal FOV, Render Quality=1) Here not corrected by frequency 75/90 becasue larger FOV
Pimax8K    SS=38%   2346x2066x2=9.69 Mpixels (72HZ, with Normal FOV, Render Quality=1)
Pimax5K+  SS=38%   2375x2029x2=9.63 Mpixels (72HZ, with Normal FOV, Render Quality=1)
Pimax5KSuper  SS=60%   2368x2014x2=9.53 Mpixels (90HZ, with Normal FOV, Render Quality=1)
VivePro     SS=106%  2076x2306x2=9.57 Mpixels 
Odyssey    SS=188%  1958x2436x2=9.54 Mpixels
old WMR   SS=188%  2184x2184x2=9.54 Mpixels
Rift CV1    PD=1.49   1882x2240x2=9.55 Mpixles (222% SS)
Vive           SS=188%  2073x2303x2=9.55 Mpixels

 

Turn-off any kind of ASW (also called Motion Smoothing in Steam VR, Motion Reprojection in WMR, etc).

Also, be sure that "Legacy Reprojection" in the per Application setting of SteamVR for IL-2 is DISABLED. (Legacy Reprojection is the old Motion Smoothing technique)

Turn-off the advanced sumpersample filtering of SteamVR.

Set IL-2 Graphics option exactly the same than the ones used in the CPU test above, but enabling VR mode.

 

Go to Missions, set Realism to Normal and select BenchmarkVx mission and Start button. Then the map view appears and press Start button. 

Wait about 25 seconds (load buildings)

Press "M" to hide the Map.

Press "P" and don´t move your head while the mission run.

 

9. VR TEST2 (OPTIONAL)

 

Use the same IL-2 settings that VRTest1, but set SteamVR %SS to a value to reach around 19.5 Million pixels which is the common basis for high-res (based on Reverb G2 at 100%)

This test is intended to stress more the GPU which is the use case for users with high resolution VR devices with SS closer to the recommed by the VR vendor.

 

Reverb G2  SS=100%    3162x3093x2=19.56 Mpixels 

Index          SS=216%   2963x3292x2=19.51 Mpixels
RiftS           PD=1.93   3187x3435x2=21.9 Mpixels at 80Hz (374% SS)

Reverb G1  SS=206%   3165x3100x2=19.62 Mpixels 

Varjo XR3   SS=58%     3119x3119x2=19.46 Mpixels

VivePro2     Set Ultra mode (2448x2448@90Hz) and set SS to have around 3128x3128 pixels per eye.

Varjo Aero SS=116%  3382x2899=19.61 Mpixels with High Settings (default) in Varjo base software

Pico4        SS=130% 3117x3117=19.43Mpixels (remember to use 90Hz mode)

Quest3      SS=260% 3016x3248x2=19.59Mpixels (remember to use 90Hz mode and 1.0 custom resolution in Oculus)

Crystal      SS=44% 2860x3383x2=19.35Mpixels (use 90Hz and custom resolution=1.0)

BSB1&2     SS=138% 3134x3134x2=19.64Mpixels (use 90Hz)

Pimax8KX  SS=83%   3384x2876x2=19.5 Mpixels (75HZ, with Normal FOV, Render Quality=1)
Pimax8K    SS=56%   3317x2922x2=19.39 Mpixels (72HZ, with Normal FOV, Render Quality=1)
Pimax5K+  SS=78%   3402x2907x2=19.78 Mpixels (72HZ, with Normal FOV, Render Quality=1)
Pimax5KSuper  SS=123%   3390x2884x2=19.55 Mpixels (90HZ, with Normal FOV, Render Quality=1)

 

10. VR TEST3 (OPTIONAL)

Use the same IL-2 settings that VRTest2, but adding Clouds to Extreme and MSAAx8

This test is only designed to test VR in maximum load of both CPU and GPU.

 

 

Note: If the original author wants to re-create his thread, I will delete this and pin the author's one! 

 

 

[chiliwili69]

On 11/24/2025 at 10:09 PM, LuftManu said:

Note: If the original author wants to re-create his thread, I will delete this and pin the author's one! 

Thanks for taking the iniciative here.

I think this test should be just left to test IL-2 GB (and perhaps a new benchmark for Korea) for CPUs and GPU in monitor.

Testing VR performance is not an easy thing (it really requires fpsVR) and it depends on many specific parameters like FOVs, frequencies or features like FFR, DFR , upscaling, third party tools and lately Foveated Encoding.

So, perhaps we should remove the VR tests since it will be difficult to properly compare apples with apples.

[LuftManu]

13 hours ago, chiliwili69 said:

Thanks for taking the iniciative here.

I think this test should be just left to test IL-2 GB (and perhaps a new benchmark for Korea) for CPUs and GPU in monitor.

Testing VR performance is not an easy thing (it really requires fpsVR) and it depends on many specific parameters like FOVs, frequencies or features like FFR, DFR , upscaling, third party tools and lately Foveated Encoding.

So, perhaps we should remove the VR tests since it will be difficult to properly compare apples with apples.

Thanks to you Chiliwilli! 

Testing is not easy at all as you said! But your chart is really useful, you found out some "comparision" between Firestrike Ultra and Digital Nature. Let's see if the thing is the same in Korea engine! 

Maybe some baseline on 4K can then help into VR.

Kind regards,

[RavenSystem]

VR is very different. Rendering an image of 2000x1000 for a flat display is not the same as rendering two images (one per eye) of 1000x1000 for VR. The CPU must do about the double of works because it must render 2 images (2 eyes) for each frame instead one. And there are other technical details that affect to VR, for example, the get_pose functions, the barrel distortions, the HMD display type (direct or indirect), etc...

[LuftManu]

17 minutes ago, RavenSystem said:

VR is very different. Rendering an image of 2000x1000 for a flat display is not the same as rendering two images (one per eye) of 1000x1000 for VR. The CPU must do about the double of works because it must render 2 images (2 eyes) for each frame instead one. And there are other technical details that affect to VR, for example, the get_pose functions, the barrel distortions, the HMD display type (direct or indirect), etc...

Hi, Raven.

Thanks for your input!

Then let's keep this for Screen performance testing only then!