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

[RAAF492SQNOz_Steve]

52 minutes ago, J2_SteveF said:

I let the hype of the Zen 3 get the better of me and ended up with a 5600 on a gigabyte board.

 

TBH I'm not overwhelmed with the Ryzen. It's a definite performer out of the box, but does not have a lot of head room for OC. In fact I've ended up leaving it on PBO. I also used a gigabyte board for the first time  in 20 + years of building my own PCs.

I'm not overly impressed with that either, got an Aorus X570 Elite. Had to flash the bios to be able to boot with the 5600 on first start up. Which although not ideal I could get past.  But then then had an intermittent power off issue for the first few weeks. PC would just power off randomly. Had it going 48 hours on Prime 95 with no issue. But then after 10 mins browsing the web the  PC powered off. Issue seems to be resolved now with the latest bios that was released a few weeks ago. So far Gigabyte BIOS seems far from polished.

 

I was going to try and get a 6800xt to go with this set up , but have now decided to stick with Nvidia and hold on for.a 3080

 

 

 

 

Think it is a good idea to go for a NVidia GPU at present.

 

After also being sucked in by the Zen 3 hype, I focussed on pairing it with RAM that had good tuning capability. Are you able to get the 5600X to run at 3800 MHz RAM frequency with FLCK 1900 etc?  While there does not seem to be much benefit in OC'ing the CPU (In my very very limited experience) the Ryzen does respond well to good memory tuning and tight timings.

 

I should also mention that I am pleased that I did the switch to the, in my case to the 5800x, AMD CPU. Gained about a 15% improvement after the switch from a i7 10700kf.

Edited February 2, 2021 by RAAF492SQNOz_Steve

[HunDread]

22 hours ago, J2_SteveF said:

TBH I'm not overwhelmed with the Ryzen. It's a definite performer out of the box, but does not have a lot of head room for OC

 

In my view that's the good thing about Ryzen. You put it in and it works (not considering your MB issues now). No real necessity to fiddle around and study the overclocking guides like I had to do with Intel. And thinking this one step further the whole strategy of intel's -K versions is like "we can produce good CPU's but do not guarantee you anything. If you end up with base clocks and can't go further we are not responsible for anything"

 

People including myself were taking OC'ing a natural thing and were so happy to squeeze out a little bit more juice we felt we managed to beat the system while in reality the system beat us by silently placing all the responsibility on us, the customers.

[J2_Steve]

yI do agree to a certain point with the above, but I have always been a fella that buys below high end and clock to high end speed. The issue I have is PBO is on by default for the Ryzens. My previous CPU MB was a X299 and 7740x @5.0, which was a £400 set up, givingt me a Single thread passmark score of 3330. I've now spent £500 on a Ryzen 5600 and cant get anything above PBO and Passmark and a single thread score of 3400. Although I haven't benched using Vanders mission. I have not noticed any real improvement in performance out of IL2 from the new cpu. No doubt there is improvement.

 

So while it's great that they are fairly quick out of the box. We are paying a premium for it. 

 

And right now I'm fairly pissed off with NVidia and AMD for their paper releases, releasing a few underpriced units to drum up interest. Then not being able to produce any more units, which ironically are now no where near the press release prices.

 

 

 

Edited February 5, 2021 by J2_SteveF

[chiliwili69]

On 2/5/2021 at 2:31 PM, J2_SteveF said:

My previous CPU MB was a X299 and 7740x @5.0, which was a £400 set up, givingt me a Single thread passmark score of 3330

3330 is a pretty high mark for an Intel CPU. I wonder how fast will run the Syn_Vander.

[Jaws2002]

I reinstalled windows and converted the file system from MBR to GTP, so I gave it another go at the benchmark. For some reason my infinity fabric won't go to 1900Mhz with this new bios, so I just used straight 3600Mhz XMP ( 16-16-16-36)  and no overclocking at all for now. 

    Solid boost in performance and general responsiveness, because of fresh Windows and finally got around to switch my BIOS to UEFI.

 

 

Motherboard:        Gigabyte X570 Aorus Master

CPU:                        Ryzen 9 5950X, (out of the box, fresh windows install).

CPU frequency:     5.05GHz

L3 Cache:              64 MB

Cores                     16

Threads                 32

RAM type             DDR4 

Ram size              32GB  (2x16GB)

Ram frequency   3600MHz

CAS                      16    timings 16-16-16-36 Aida64 latency 62ns

NB frequency    1800MHz

GPU                      Gigabyte Aorus 1080Ti (factory overclocked)

 

1080p CPU test.

2021-02-06 23:09:03 - Il-2
Frames: 8135 - Time: 60000ms - Avg: 135.583 - Min: 120 - Max: 189

2021-02-06 23:10:39 - Il-2
Frames: 8090 - Time: 60000ms - Avg: 134.833 - Min: 120 - Max: 186
 

[chiliwili69]

4 hours ago, Jaws2002 said:

Frames: 8135 - Time: 60000ms - Avg: 135.583 - Min: 120 - Max: 189

 

Thanks for this. You improved your previous results with FCLK 1900, that´s interesting.

In theory, the disk thing shold not influence the performance since everything is between CPU, RAM, GPU. Perhaps the fresh windows install eliminated some software or processes which were affecting your previous performance.

 

I wonder now how far you can go now with the new ClockTuner tweaking and/or with some RAM timming tweaking...

[NiiranenVR]

A little bit out of  sentence -  well , part of the hobby to play the game is allso to have the right stof in pc so you can play the game .......

So ... I have a ssd hd , win 10  on one and il2 on 2'cond - ............and as I'm understanding it could give no nano better performance to change to ssd M2 socket / high speed ??? ?

[Wilhelm_Heidkamp]

5 hours ago, J5_NiiranenVR-Gfr said:

A little bit out of  sentence -  well , part of the hobby to play the game is allso to have the right stof in pc so you can play the game .......

So ... I have a ssd hd , win 10  on one and il2 on 2'cond - ............and as I'm understanding it could give no nano better performance to change to ssd M2 socket / high speed ??? ?

 

In January I changed my SSD by an AORUS NMVE M2 2TB disk. 2500MB/s. I haven't seen better performance in game because that.

Edited February 7, 2021 by PA_Willy

[Jaws2002]

7 hours ago, chiliwili69 said:

 

Thanks for this. You improved your previous results with FCLK 1900, that´s interesting.

In theory, the disk thing shold not influence the performance since everything is between CPU, RAM, GPU. Perhaps the fresh windows install eliminated some software or processes which were affecting your previous performance.

 

I wonder now how far you can go now with the new ClockTuner tweaking and/or with some RAM timming tweaking...

 

I was surprised as well. I had to go through the settings to make sure i didn't miss something. Everything looks the same as before. In nvidia control panel i have all default except power management to max performance and Vsinc to OFF. In game it's exactly like your screenshot, but i clicked "full screen mode". My monitor is 1440p and does some weird things in windowed mode.

   One other difference from my previous windows install is that i got rid of Norton 360. Now i use just Windows defender.

 

 I gave it another go this morning, just to be sure that run was not a fluke.

 

2021-02-07 11:28:30 - Il-2
Frames: 8114 - Time: 60000ms - Avg: 135.233 - Min: 117 - Max: 181

2021-02-07 11:58:37 - Il-2
Frames: 8092 - Time: 60000ms - Avg: 134.867 - Min: 116 - Max: 185

2021-02-07 12:00:09 - Il-2
Frames: 8021 - Time: 60000ms - Avg: 133.683 - Min: 115 - Max: 177

 

 

13 minutes ago, PA_Willy said:

 

In January I changed my SSD by an AORUS NMVE M2 2TB disk. 2500MB/s. I haven't seen better performance in game because that.

 

 Same here. The big boost in performance is going from spinning drive to SSD. Going from Sata SSD to M.2 NVMe SSD  is not noticeable in game.

It makes a huge difference when you have to transfer large files.

Edited February 7, 2021 by Jaws2002

[RAAF492SQNOz_Steve]

On 2/1/2021 at 7:20 PM, chiliwili69 said:

 

Many thanks for this detailed test. It gives a lot of data. And the top record in the VRtest2!

Just for completeness it would be good if you can run just the CPUtest for the Ryzen and 3800MHz.

 

It is also interesting that the FLCK 1900 vs 2000 give same results.

 

Regarding the reprojection (motion smoothing), your minimun fps is not 45, so I don´t know how fpsVR was detecting that.

 

 

 

Re: FLCK 1900 vs 2000

 

Re-ran the SYN VANDER Benchmark VR2 test setup after I had tightened primary and secondary memory timings with intermediate operating frequencies as well to see if their was a "sweet spot" frequency between 3800 - 4000MHz.

Results were pretty similar with 3800 MHz giving a better result than 4000 MHz RAM again.  Looks like I will be leaving my RAM at 3867 MHz although if you can run at 3800 MHz you will get pretty similar results.

 

Tests conducted 08/02/21, ambient room temp 20 C. Three tests run for each frequency and averaged result in bold.
                                
VR2 Test AMD 16-15-15-15-30 4000MHz
    Frames   Min    Max    Average        Latency
    4880        60    91    81.333        55.2 ns
    4819        59    91    80.317        
    4879        60    91    81.317        
    4859.3333333333        59.6666666667   91       80.989        
                                                        
VR2 Test AMD 16-15-15-15-30 3933MHz

Frames   Min    Max    Average        Latency
    4688        57    91    78.133        55.9 ns
    4843        60    91    80.717        
    4874        58    91    81.233        
    4801.6666666667        58.3333333333   91       80.0276666667        
                                              
VR2 Test AMD 16-15-15-15-30 3867MHz

Frames   Min    Max    Average        Latency
    4980        63    91    83        56.6 ns
    4958        63    91    82.633        
    4821        58    91    80.35        
    4919.6666666667        61.3333333333   91       81.9943333333        
                                                     
VR2 Test AMD 16-15-15-15-30 3800MHz

Frames   Min    Max    Average        Latency
    4841        58    91    80.683        57.1 ns
    4820        59    91    80.333        
    5013        62    91    83.55        
    4891.3333333333        59.6666666667   91       81.522        

 

PC Build

Ryzen 5800x. 

MSI Tomahawk X570

MSI Trio RTX 3090

Patriot Viper 4400 MHz, 2 x 8 Gb.

Samsung 1 Tb M.2 NVME storage

HP Reverb G2

 

Think if I run the SYN VANDER Benchmark any more I will come out in a rash!?

Edited February 8, 2021 by RAAF492SQNOz_Steve
clarification, tidy up for pasted stuff.

[shirazjohn]

Hi guys I have recently installed better quality ram on my pc (TeamGroup DDR4 3600 CL 14-15-15-35 1.45v 2x 8 Gb samsung b die.)

 

I have spent the last week tuning to try and gain the best performance, I am a novice at this so used Ryzen dram calculator as a starting point and used the following guide for help.

 

https://github.com/integralfx/MemTestHelper/blob/master/DDR4 OC Guide.md

 

I am quite pleased with the results, a good improvement over previous results, especially the vr tests (not bad for an old gtx 1080ti running a reverb g2).

 

Cpu: R5600x

Cpu Freq: 4.8 Ghz

L3 Cache: 32 Mb

Cores: 6

Threads: 12

Ram type: ddr4

Ram Size: 16Gb (2x8gb)

Infinity clk: 2000 Mhz

Ram Freq: 4000 Mhz

Ram Timings: 16-16-16-30-48-trfc320

Gpu: Gtx 1080ti core + 165 Mhz / Mem +480 Mhz

Aida Latency: 54.6 ns

 

ddr 4000/16-16-16-30-48-trfc 320/pbo + 150 mhz/-5 all core curve optimer/ pbo limits= motherboard 

Cpu Max temp 50 degrees during test 

 

2021-02-11 19:42:53 - Il-2
Frames: 7827 - Time: 60000ms - Avg: 130.450 - Min: 116 - Max: 172

2021-02-11 19:44:55 - Il-2
Frames: 7723 - Time: 60000ms - Avg: 128.717 - Min: 115 - Max: 170

2021-02-11 19:46:45 - Il-2
Frames: 7565 - Time: 60000ms - Avg: 126.083 - Min: 111 - Max: 161

2021-02-11 19:48:33 - Il-2
Frames: 7796 - Time: 60000ms - Avg: 129.933 - Min: 115 - Max: 171

2021-02-11 19:50:25 - Il-2
Frames: 7643 - Time: 60000ms - Avg: 127.383 - Min: 110 - Max: 167
                 
                         5 run Avg: 128.51         113.4      168.20

-------------------------------------------------------------------------
Il2 Vr test 100% Reverb G2 / 1080ti / Max temp 45 degrees

 

2021-02-11 20:09:54 - Il-2
Frames: 4073 - Time: 60000ms - Avg: 67.883 - Min: 57 - Max: 81

2021-02-11 20:11:39 - Il-2
Frames: 4156 - Time: 60000ms - Avg: 69.267 - Min: 59 - Max: 85

2021-02-11 20:13:26 - Il-2
Frames: 4121 - Time: 60000ms - Avg: 68.683 - Min: 59 - Max: 80

--------------------------------------------------------------------------
Il2 Vr test 50% Reverb G2 1080ti 

 

2021-02-11 20:21:03 - Il-2
Frames: 5294 - Time: 60000ms - Avg: 88.233 - Min: 76 - Max: 92

2021-02-11 20:22:48 - Il-2
Frames: 5167 - Time: 60000ms - Avg: 86.117 - Min: 69 - Max: 91

2021-02-11 20:24:36 - Il-2
Frames: 5220 - Time: 60000ms - Avg: 87.000 - Min: 70 - Max: 92

 

Edited February 14, 2021 by shirazjohn

[chiliwili69]

On 2/8/2021 at 4:45 AM, RAAF492SQNOz_Steve said:

Results were pretty similar with 3800 MHz giving a better result than 4000 MHz RAM again.  Looks like I will be leaving my RAM at 3867 MHz although if you can run at 3800 MHz you will get pretty similar results.

 

Thank you for this. It seems that in the VRtest2 the impact of the RAM is minimal. There is a variable +-2fps noise probably due to the default AMD PBO or the GPU overclock.

It could be also that at this really high rate (80s fps) you are bounded by GPU in this 19.5Mpixel test. Or by any bus bandwidth.

 

I wonder how this memory freq will affect the VRtest1 or the CPU test. But don´t need to run more test, you already did a lot.

7 hours ago, shirazjohn said:

5 run Avg: 128.51         113.4      168.20

 

Thanks for this. You got some extra fps by tunning the memory. And achieved with just 2x8 Gb sticks (not Dual rank) results on the 5900X league. Nice!

 

Thank you also for the guide. One day I have to study that matter as well and see if I can get some extra fps.

[shirazjohn]

50 minutes ago, chiliwili69 said:

Thanks for this. You got some extra fps by tunning the memory. And achieved with just 2x8 Gb sticks (not Dual rank) results on the 5900X league. Nice!

 

Thank you also for the guide. One day I have to study that matter as well and see if I can get some extra fps.

 

Hi Chilli tighter memory timings are definitely the way to go, its a bit time consuming but worth it. 

[HunDread]

Did some tests over the weekend.  I was curious about the impact of some of the in-game graphics settings compared to my default settings I use in VR. The tests were done in 4k, not in VR.

 

Conclusions:

2X FXAA              (-4 FPS over No AA)

4x FXAA              (-4 FPS over No AA) - 2x and 4x FXAA is exactly the same

2x MSAA             (-25 FPS over No AA) - MSAA is FPS Killer

Clouds - Extreme               (-12 FPS over High Clouds)

Distant Landscape Details - 4x             (-0/1 FPS over Distant Landscape 3x)

Horizon Draw Distance - 150 km           (-1 FPS over Horizon distance 100 km)

Distant Buildings - ON            (No impact)

Combined Distant Landscape Details - 4x / Horizon Draw Distance - 150 km / Distant Buildings - ON            (-3 FPS over 3x / 100 km / Off)

 

 

Details:

 

Hardware:

CPU            5600x
GPU            1080TI
RAM            3600 CL16

Screen        Flat screen

--------------------------------------

Default settings:

 

Resolution:         3840x2160 (4k)
Preset            Balanced
Shadows            Off
Mirrors            Off
Distant landscape    3x
Canopy reflections    Off
Horizon draw distance    100km
Landscape filter    Blurred    
Terrain roughness    Low
Grass quality        Normal
Clouds quality        High
AA            NO
Full screen        NO
SSAO            NO
HDR            Yes
Sharpen            Yes
4k textures        Yes
Distant buildings    NO

-----------------------------------------------------------------

Default

 

Frames: 7309 - Time: 60000ms - Avg: 121.817 - Min: 98 - Max: 135
Frames: 7407 - Time: 60000ms - Avg: 123.450 - Min: 102 - Max: 139
Frames: 7346 - Time: 60000ms - Avg: 122.433 - Min: 101 - Max: 137
Frames: 7379 - Time: 60000ms - Avg: 122.983 - Min: 99 - Max: 137
Frames: 7377 - Time: 60000ms - Avg: 122.950 - Min: 101 - Max: 137
Frames: 7374 - Time: 60000ms - Avg: 122.900 - Min: 101 - Max: 137

-----------------------------------------------------------------

2X FXAA (-4 FPS over No AA)

 

Frames: 7118 - Time: 60000ms - Avg: 118.633 - Min: 98 - Max: 134
Frames: 7124 - Time: 60000ms - Avg: 118.733 - Min: 98 - Max: 132
Frames: 7086 - Time: 60000ms - Avg: 118.100 - Min: 97 - Max: 132

-----------------------------------------------------------------

4x FXAA (-4 FPS over No AA)

 

Frames: 7121 - Time: 60000ms - Avg: 118.683 - Min: 98 - Max: 132
Frames: 7135 - Time: 60000ms - Avg: 118.917 - Min: 99 - Max: 133
Frames: 7132 - Time: 60000ms - Avg: 118.867 - Min: 99 - Max: 132

----------------------------------------------------------------

2x MSAA (-25 FPS over No AA)

Frames: 5850 - Time: 60000ms - Avg: 97.500 - Min: 83 - Max: 108
Frames: 5864 - Time: 60000ms - Avg: 97.733 - Min: 86 - Max: 108
Frames: 5866 - Time: 60000ms - Avg: 97.767 - Min: 85 - Max: 108

----------------------------------------------------------------

Clouds - Extreme (-12 FPS over High Clouds)

 

Frames: 6659 - Time: 60000ms - Avg: 110.983 - Min: 92 - Max: 124
Frames: 6650 - Time: 60000ms - Avg: 110.833 - Min: 93 - Max: 122
Frames: 6686 - Time: 60000ms - Avg: 111.433 - Min: 94 - Max: 124

----------------------------------------------------------------

Distant Landscape Details - 4x (-0/1 FPS over Distant Landscape 3x)

Frames: 7327 - Time: 60000ms - Avg: 122.117 - Min: 102 - Max: 137
Frames: 7261 - Time: 60000ms - Avg: 121.017 - Min: 101 - Max: 134

-----------------------------------------------------------------

Horizon Draw Distance - 150 km (-1 FPS over Horizon distance 100 km)

 

Frames: 7261 - Time: 60000ms - Avg: 121.017 - Min: 99 - Max: 135
Frames: 7268 - Time: 60000ms - Avg: 121.133 - Min: 100 - Max: 136
Frames: 7302 - Time: 60000ms - Avg: 121.700 - Min: 98 - Max: 135

-----------------------------------------------------------------

Distant Buildings - ON (No impact)

 

Frames: 7364 - Time: 60000ms - Avg: 122.733 - Min: 99 - Max: 137
Frames: 7414 - Time: 60000ms - Avg: 123.567 - Min: 101 - Max: 138
Frames: 7372 - Time: 60000ms - Avg: 122.867 - Min: 101 - Max: 136
Frames: 7364 - Time: 60000ms - Avg: 122.733 - Min: 101 - Max: 136

-----------------------------------------------------------------

Distant Landscape Details - 4x
Horizon Draw Distance - 150 km
Distant Buildings - ON

 

(-3 FPS over 3x / 100 km / Off)

 

Frames: 7131 - Time: 60000ms - Avg: 118.850 - Min: 97 - Max: 132
Frames: 7149 - Time: 60000ms - Avg: 119.150 - Min: 98 - Max: 134
Frames: 7151 - Time: 60000ms - Avg: 119.183 - Min: 100 - Max: 132

-----------------------------------------------------------------

[Gambit21]

On 2/2/2021 at 12:56 PM, J2_SteveF said:

I let the hype of the Zen 3 get the better of me and ended up with a 5600 on a gigabyte board.

 

TBH I'm not overwhelmed with the Ryzen. It's a definite performer out of the box, but does not have a lot of head room for OC. 

 

 

Would you rather they’d have de-tuned it so that you could overclock it to bring it up to it's current out of the box performance? The whole “I need to overclock” thing is silly. Much better if the chip is optimized to begin with thus negating the need to overclock.

[LuftManu]

2 hours ago, Gambit21 said:

 

Would you rather they’d have de-tuned it so that you could overclock it to bring it up to it's current out of the box performance? The whole “I need to overclock” thing is silly. Much better if the chip is optimized to begin with thus negating the need to overclock.

I guess it's like the  $.99. People buy what seems to them "bigger and more powerful." Old mind PR tricks. If this thing OC it's better than this other one, even if the latter is already working ?

[Dark_Jak]

Hello guys, just finish put my "new" rig online. Haven't play the game yet but manage to do the benchmarks 

Motherboard: Asus B550F-Gaming
 CPU: AMD 5600X  (Stock)            
 CPU Freq: 4.6GHz
 L3 cache:  32MB
 Cores: 4
 Threads:  12
 RAM type: Patriot Viper DDR4 4000mhz 
 RAM size: 16Gb (2x8Gb)
 RAM Freq: 1999.5Mhz
 RAM timings: 19-19-19-39 (D.O.C.P)
 FCLK: 2000
 GPU: Gigabyte RTX 2070 Super Gaming OC 3X
 Aida64 Timings (62.1ns)

CPU Test:
Frames: 6363 - Time: 60000ms - Avg: 106.050 - Min: 93 - Max: 142

VR Test: Oculus Rift S PD:1.36
Frames: 4410 - Time: 60000ms - Avg: 73.500 - Min: 38 - Max: 81

I'm still using the D.o.c.p profile of the ram stick. Still need to tight the timings a little bit more. 

 

Edited February 22, 2021 by Dark_Jak

[chiliwili69]

7 hours ago, Dark_Jak said:

Haven't play the game yet but manage to do the benchmarks

 

Thank you Dark_Jak for you result and congrats for your new rig.

 

I think it is the first test with the new 4.506 update. Your 106.0 value is a bit below expected. I don´t know if this is because the new release or because you RAM freq-timing.

You achieve 2000 FCLK but using 19-19-19-39.

Perhaps you can try to lower the FCLK to for example 1800 or 1900 and decrease the timming. Trying other xmp/docp profiles.

Look at RAAF492SQNOz_Steve test results with Viper RAM as well

[Dark_Jak]

9 hours ago, chiliwili69 said:

 

Thank you Dark_Jak for you result and congrats for your new rig.

 

I think it is the first test with the new 4.506 update. Your 106.0 value is a bit below expected. I don´t know if this is because the new release or because you RAM freq-timing.

You achieve 2000 FCLK but using 19-19-19-39.

Perhaps you can try to lower the FCLK to for example 1800 or 1900 and decrease the timming. Trying other xmp/docp profiles.

Look at RAAF492SQNOz_Steve test results with Viper RAM as well

 

Agree with you chili, wasn't expecting that "low" performance in the CPU test. 

I only have 2 xmp timings available 4000 cl19 and 3866 cl18, does being automatic xmp profiles. 

Later tonight I'm gonna do some more test. 

The best I manage to get so far without crash was 3600mhz cl15-15-15-30 at 1.37v

Also RAAF492SQNOz_Steve is running the 4400mhz kit wich I think its easy to downclock to 3800 cl14 

[Dark_Jak]

Here are the the only 2 test I manage to do.

3600mhz cl15-15-15-32 (Only change the primary timings) 

Frames: 6891 - Time: 60000ms - Avg: 114.850 - Min: 93 - Max: 153

Frames: 7132 - Time: 60000ms - Avg: 118.867 - Min: 106 - Max: 158

Frames: 7131 - Time: 60000ms - Avg: 118.850 - Min: 105 - Max: 154

 

3800mhz CL17-17-17-36 

Frames: 6998 - Time: 60000ms - Avg: 116.633 - Min: 101 - Max: 156

Frames: 7202 - Time: 60000ms - Avg: 120.033 - Min: 99 - Max: 158

Frames: 6569 - Time: 60000ms - Avg: 109.483 - Min: 89 - Max: 149

Don't know if does are around a good average for the 5600X. 
Still don't know if I send back this Patriot and get some Ballistix 3600mhz Cl16

 

[chiliwili69]

7 hours ago, Dark_Jak said:

Frames: 7131 - Time: 60000ms - Avg: 118.850 - Min: 105 - Max: 154

 

This is a much better result. Well aligned with all the other tests for your CPU.

These tests also confirms that going to lower freqs with lower timmings can be better.

This also confirms that the new game update doesn´t affect performance at all.

 

I don´t know how far you can go with better RAM, you are already well above of most of the 5600X tested.

[HunDread]

14 hours ago, chiliwili69 said:

This also confirms that the new game update doesn´t affect performance at all.

 

I did test this when the patch came out. No real difference indeed:

 

Last test a day before patch (settings do not really matter but they are a few posts up)

 

Frames: 7327 - Time: 60000ms - Avg: 122.117 - Min: 100 - Max: 137
Frames: 7367 - Time: 60000ms - Avg: 122.783 - Min: 102 - Max: 138
Frames: 7358 - Time: 60000ms - Avg: 122.633 - Min: 101 - Max: 136

 

 

Test after new patch (same settings)

 

Frames: 7287 - Time: 60000ms - Avg: 121.450 - Min: 99 - Max: 136
Frames: 7316 - Time: 60000ms - Avg: 121.933 - Min: 102 - Max: 135
Frames: 7301 - Time: 60000ms - Avg: 121.683 - Min: 101 - Max: 135

[chiliwili69]

9 hours ago, HunDread said:

I did test this when the patch came out. No real difference indeed

Thanks for these tests. So we can mix the results with previous test.

[Dark_Jak]

Update: 

So decide to try some test with the memory at 3733mhz cl16-17-17-34 trfc-360 only changed the primary timings to see how it works, cause I was having some troubles with them running at 3800mhz

This was also the first time I've done the VR Test with the correct settings the last one I thought I've changed the PD to 1.36 but It didnt. So here are the results:

Stock Settings on CPU
RAM: 3733mhz CL-16-17-17-34 trfc-360 DramV-1.40
CPU Test: 1080p

Frames: 7040 - Time: 60000ms - Avg: 117.333 - Min: 102 - Max: 155

Frames: 7219 - Time: 60000ms - Avg: 120.317 - Min: 104 - Max: 161

Frames: 7209 - Time: 60000ms - Avg: 120.150 - Min: 105 - Max: 163

VR Test
Oculus Rift S - OpenCoposite - OTT: PD:1.36
 

Frames: 3834 - Time: 60000ms - Avg: 63.900 - Min: 38 - Max: 81

Frames: 3921 - Time: 60000ms - Avg: 65.350 - Min: 38 - Max: 81

Frames: 3924 - Time: 60000ms - Avg: 65.400 - Min: 38 - Max: 81

One question, will it improve if I active PBO in games in general or should I leave it off, or go with manual O.C?

Edited February 28, 2021 by Dark_Jak

[chiliwili69]

19 hours ago, Dark_Jak said:

Frames: 3924 - Time: 60000ms - Avg: 65.400 - Min: 38 - Max: 81

One question, will it improve if I active PBO in games in general or should I leave it off, or go with manual O.C?

 

Thank you for your test.

 

Your VR test1 with 1.36PD seems that it is a bit compromised for your 2070S, but normally with the RiftS having 1.1 or 1.2 is quite OK. Your RiftS is well complemented by the 2070S.

 

Regarding the PBO, we still don´t know to what extent it helps to the performance. Currently I don´t use it but perhaps it could help. I really don´t know.

 

[Dark_Jak]

On 2/28/2021 at 8:26 PM, chiliwili69 said:

 

Thank you for your test.

 

Your VR test1 with 1.36PD seems that it is a bit compromised for your 2070S, but normally with the RiftS having 1.1 or 1.2 is quite OK. Your RiftS is well complemented by the 2070S.

 

Regarding the PBO, we still don´t know to what extent it helps to the performance. Currently I don´t use it but perhaps it could help. I really don´t know.

 

Yeah, I know the 2070S ain't as good for VR as 2080 TI or 3080. I will do one more test performance but with the PD ate stock 1.0. I was using it before I made the upgrade for the 5600X and was still playable with some stutter from time to time. 

Just done some more test this time also in VR with some thighther timings on the ram and also O.C my 2070S. 

Here the results 

1080P CPU Test Settings  - 3733mhz CL16-16-16-32 trfc320 -  RTX 2070S +85CoreClk +550MemoryClk

Frames: 6800 - Time: 60000ms - Avg: 113.333 - Min: 98 - Max: 147

Frames: 7186 - Time: 60000ms - Avg: 119.767 - Min: 105 - Max: 161

Frames: 6856 - Time: 60000ms - Avg: 114.267 - Min: 100 - Max: 152


VR Test1 - Rift S  1.37PD RTX 2070S +85CoreClk +550MemoryClk 


Frames: 4595 - Time: 60000ms - Avg: 76.583 - Min: 60 - Max: 81

Frames: 4640 - Time: 60000ms - Avg: 77.333 - Min: 53 - Max: 81

Frames: 4651 - Time: 60000ms - Avg: 77.517 - Min: 66 - Max: 81
 

 

 

Edited March 8, 2021 by Dark_Jak
Update on the system so far.

[RufusK]

Motherboard:   ASUS TUF Gaming X570
 CPU:                5800X
 CPU Freq:      4.49 Ghz
 L3 cache:      32 MB
 Cores:           8 
 Threads:       16 
 RAM type:     DDR4
 RAM size:       32Gb (2x16GB)
 NB Freq:        1796.4 MHz 
 DRAM Freq:   1796.4 MHz 
 RAM timings:     20-20-20-40-85

 

MCLK:      1800
FCLK:       1800
UCLK:      1800

 

GPU:        6900XT

 

CPU test
Frames 5701 - Avg. 95.0 - Min. 81 - Max. 137

 

GPU test
Frames 4560 - Avg. 76.0 - Min. 65 - Max. 90

 

These results are pretty disappointing. Do I need to get better RAM or improve the timings? Or is this simply the 6000-series curse?
I have no experience with RAM overclocking, so any advice would be appreciated.

[chiliwili69]

10 hours ago, RufusK said:

CPU test
Frames 5701 - Avg. 95.0 - Min. 81 - Max. 137

 

Thank you for this test. It is the first with a 6900XT.

It also confirms the same results we had with other AMD GPUs (6800 and 6800XT). There is a issue between those cards and IL-2 game.

Two months ago I reported that to developers here

Problem could be that developers have no access to this latest AMD cards.

In other games those cards run OK as far as I know.

Maybe you can create a specific post in the bug reports thread.

[Jaws2002]

Thank you for your test RufusK. 

 I think you could do a lot better if you improve the timings of your memory, but a trend emerges with the 6000 series and this game. 

  I don't blame the cards.

 The problem is between AMD drivers and the game engine. 

 You should start a bug report thread to let the developers know they have issues with this "new"  AMD Cards. 

 While hard to find, this cards have been released four months ago, so the team should have fixed this issues by now. 

 There's no excuse to have 5000 series Cards run faster than the new 6000 series.

Edited March 12, 2021 by Jaws2002

[RufusK]

I created a post in the technical issues and bug report forum as suggested. I hope the devs are able to address this issue.

 

[vwillo56]

For us commoners.

Reverb G2
I7-10700, non OC

2060S 8G

16G 2666M 

SSD, clean & optimized current Win 10 and il-2 GB

nvidia, 3D, Global- max power, vsync-off, Program- il2, BoS

WMR for Steam - Opt. For Perf. 
SteamVR, General- 90hz, custom, RPE 2228x2184=50%.
                  Video- Per-App.- WMR for SteamVR,  Memory     Smooth- force always on, Legacy-off

IL2- High

SYN_Vander Benchmark 

VR1- BEST I could achieve, from Low thru High cycling, all parameter variations

Frames: 2691-2696, 60Kms, min-42, max-50, avg-44.86.

Surprisingly, LOW only achieved 2697, 43, 48, 44.9 and didn’t look as good.

In a quick mission. Tree tops to 2km, min-42 max-91, avg-49 over a 60 second period and was pretty acceptable. A few jaggies.

Hope this helps AND any recommendations welcomed!

Edited March 20, 2021 by vwillo56

[JG1_Vonrd]

Here are my results... both CPU and GPU recently overclocked (still adjusting to find the sweet spot):

 

 Motherboard: ASUS Maximus IX Hero
 CPU: i7 7700K 
 CPU Freq:        4.2 Ghz
 L3 cache:        8 MB
 Cores:             4 
 Threads:           8
 RAM type:        DDR4
 RAM size:        32Gb (4x8GB)
 NB Freq:         4220 MHz (Uncore Frequency)
 RAM Freq:        2142 MHz (Dual channel)
 RAM timings:     15-15-15-36-374
 GPU:             Geforce GTX 1080Ti

 

Optional:

CPU Cooler: Corsair H100i V2 Liquid
RAM Model: G. Skil (SK Hynix) F4-3000C15-8GVRB
GPU Model: Geforce GTX 1080Ti

 

Tests:

 

In 2D Monitor Acer Predator X34

1080P settings:
Frames: 4457 - Time: 60000ms - Avg: 74.283 - Min: 64 - Max: 99

 

Reverb G2

 

VR Test 1:
Frames: 2685 - Time: 60000ms - Avg: 44.750 - Min: 40 - Max: 46

 

VR Test 2:
Frames: 2692 - Time: 60000ms - Avg: 44.867 - Min: 42 - Max: 46

 

Looking at the spreadsheet it seems that my results are consistently at the low end for both my CPU and GPU. 

I'm not trying to chase FPS but am more interested in frametime / smoothness. I'm OK with how VR is working in GB (with the exception of IDing aircraft closer in. I  don't care so much about spotting pixels at extreme range but I have to hold fire before IDing the plane until too close as opposed to 2D).

 

I know this is GB but I'm so frustrated with this new Reverb G2 in DCS that I'm ready to give it up. Any advice is greatly appreciated for both GB and DCS.

Edited March 21, 2021 by JG1_Vonrd

[chiliwili69]

On 3/19/2021 at 9:20 PM, vwillo56 said:

Reverb G2
I7-10700, non OC

2060S 8G

 

On IL-2 VR there are two factors that affect the overall performance. One is CPU/RAM and the other is GPU. They both need to be balanced with their corresponding graphics settings.

For CPU/RAM (building scene): You need to adapt the Preset (Low, Balanced, High, Ultra) and also the shadows, mirrors.

For GPU (rendering scene): You need to adjust the SS%, MSAA and the clouds.

 

Please, take the app fpsVR and you will see which is your constrain when running IL-2 VR.

 

In you case I think the 2060S would be the clear bottleneck. Try to do QMB with balanced and no clouds, no AA, no shadows, no mirrors and SS% at just 30%. The image will be less nice but perhaps you reach 90. Otherwise you will need to use Motion Smoothing and be at 45fps.

 

You can also run the CPU test, just to be sure that your CPU/RAM is delivering what it is expected.

9 hours ago, JG1_Vonrd said:

RAM Model: G. Skil (SK Hynix) F4-3000C15-8GVRB

 

Thank you for your test. Your numbers are aligned to your specs.

 

I noticed that you have that memory, so in theory you should be able to run it at 3000MHZ (1500x2). There should be an xmp profile for that. This might give you some extra fps or reduce frametimes.

9 hours ago, JG1_Vonrd said:

 CPU: i7 7700K 
 CPU Freq:        4.2 Ghz

 

Also, I believe that with your current liquid cooling you could reach 4.7 or 4.8 GHz. But I have not experience with OC the 7700K. Other people in this forum with 7700K were reaching that.

[Dutch2]

I did run the i7-7700k at 5,2Ghz (prime95) and shifted it 0,1Ghz down to be sure. The 7700k is a great overclocker, if have a decent VRM on your board. 

[JG1_Vonrd]

11 hours ago, Dutch2 said:

I did run the i7-7700k at 5,2Ghz (prime95) and shifted it 0,1Ghz down to be sure. The 7700k is a great overclocker, if have a decent VRM on your board. 

So far I have it up to 4.7 Ghz and I'm pretty sure that the ASUS Maximus IX Hero has a good VRM. What temp were you seeing at your overclock?

 

I also enabled XMP on the memory.

 

I'll do another run on the bench after I do some more tweaking.

[chiliwili69]

28 minutes ago, JG1_Vonrd said:

What temp were you seeing at your overclock?

 

IL-2 uses AVX instructions, so normally when doing overclock it is better to put AVX offset to zero.

 

When running prime95 temps could go much higher than running IL-2. I think that for prime95 temps around 80-85 celsius is OK. But later when gaming IL-2 the temps will be around 70-75 celsius. But it depends as well on your room temp.

[JG1_Vonrd]

38 minutes ago, chiliwili69 said:

 

IL-2 uses AVX instructions, so normally when doing overclock it is better to put AVX offset to zero.

 

When running prime95 temps could go much higher than running IL-2. I think that for prime95 temps around 80-85 celsius is OK. But later when gaming IL-2 the temps will be around 70-75 celsius. But it depends as well on your room temp.

Understood... I was just curious as to what temps Dutch was seeing during the overclocking benchmarking.

Actually, I'm pretty much OK with how my rig is performing in IL-2... DCS... that's a whole different matter...

[Dutch2]

1 hour ago, JG1_Vonrd said:

So far I have it up to 4.7 Ghz and I'm pretty sure that the ASUS Maximus IX Hero has a good VRM. What temp were you seeing at your overclock?

 

I also enabled XMP on the memory.

 

I'll do another run on the bench after I do some more tweaking.

I do not know the temperature I’m now at the i7-9700k/Arctic 360Freezer2, but that was always below the Intel specs. AVX indeed to zero, no HT and I did have the CPU dilidid, using the Der8auer delid tool. 

My air cooler was the https://www.overclock.net/threads/thermalright-true-spirit-140-power.1477785/ only I did exchange the single stock fan by 2 Thermalright 140 high rev in push/pull. 

PC case was complete open to eliminate every heat build up. 

If you need learn about overclocking, check always videos from experience overclocking guys who know what they are doing and use your type of mobo. I think Der8aur is one of them see for your self at:

 

 

Forget one thing about temperature, I always Google a couple of sites for the max temp. and in the Bios I set this as an alarm rate. 

[kestrel79]

Anyone try IL-2 yet with a 3060 or 3060ti yet? Curious how those run in VR. When more are available I might get one or a 3070 and try overclocking my cpu for the first time.

[TWC_Ace]

I need advice. I was due to upgrade my PC this year but miners skyrocketed the GPUs prices so its on the hold for now. Though I upgraded from 24" IPS 1200p monitor to 27" 1440p IPS freesync/gsync 160hz monitor and Im seeing improvements in other games but in IL2 nothing changed in performance area (image is nicer, sharper, more detailed, contacts in the distance more noticable etc.). I dont have microstutters in MP anymore (not in the extent I used to have) but some missions and servers are not smooth as others and specialy noticable when moving my head with ED Tracker PRO. I tried to play with camera and graphic settings as well as nvidia CP settings but didnt see much improvement. Is it possible that my PC aged so it cant cope with the game even on moderate high settings at 1440p? Or the problem is somewhere else because on some missions and servers I have very smooth gameplay, as well in SP and other games like Post Scriptum, Squad etc. FPS is OK, usualy 77-110 (I locked it to 80 now). Is it CPU or?

My rig:

i5 7600 (turbo at 3.9Ghz)

MBO: Asrock h110m dgs r3.0

GPU: Gigabyte GTX 1660 OC 6GB (~gtx1070)

RAM: 16GB (2x8) DDR4 HyperX Fury 2400Mhz

SSD

Win10 pro 20H2 64bit

Monitor: Dell S2721DGFA, 1440p, nano LG IPS panel, 1/4ms, 160Hz (DP)

Flight gear: VKB Gladiator pro, MFG Crosswind pedals, Edtracker Pro head tracking system

Edited March 28, 2021 by =VARP=Tvrdi