Jump to content
chiliwili69

Benchmark for CPU/RAM performance: Remagen 4.002 to 4.004

Recommended Posts

On 1/17/2020 at 12:23 AM, chiliwili69 said:

 

Thank you very much for bringing the first 9900KS results!! The 5.4 GHz freq is impressive as well.

 

 

 

 

Running with the same settings in VR on Reverb 2200x2152, beta versions.

 

2020-01-18 12:05:49 - Il-2
Frames: 14425 - Time: 180000ms - Avg: 80.139 - Min: 51 - Max: 91

Edited by jarg1

Share this post


Link to post
Share on other sites

Maybe it will be interesting. I did a test on 5.3 and got these results.

Frames: 33401 - Time: 192047ms - Avg: 173.921 - Min: 110 - Max: 287 - 5.3GHz

Frames: 34904 - Time: 194813ms - Avg: 179.167 - Min: 120 - Max: 287 - 5.4GHz

Changes in 100 MHz changed the minimum and average fps, but the maximum remained the same.

Share this post


Link to post
Share on other sites
13 hours ago, jarg1 said:

Frames: 14425 - Time: 180000ms - Avg: 80.139 - Min: 51 - Max: 91

 

Thanks for sharing this. You are not far from the 90. 

It is good to know that even with the most powerful machines, being at 90 full time is still an utopia. Other panel frequencies like 80Hz (Rift-S, Index) or 72Hz(Pimax) or even 60Hz(Reverb) has better chances to be there.

Edited by chiliwili69

Share this post


Link to post
Share on other sites

For people with 9900K or 9900KS chips, have you tried to disable 2 or 4 cores?

I don´t know how this will affect performance for the same freq.

Also, with less cores active, I believe the heat produced is less and the required voltage is lower. Do you think so?

Share this post


Link to post
Share on other sites
16 hours ago, chiliwili69 said:

 

I have to try that with fpsVR.

 

 

By the way I think you misinterpret the output of this tool. I can't use it because I am  a rift user but from what I quickly read your interpretation is wrong, appolon01 has the right interpretation.

 

The curve does not represent the history of a frame. The X axis does not represent what happens during a frame. It is a distribution of the frame time.

 

When the blue curve is high, say 6% at 11 ms,  it means that 6% of the frame took 11 ms. So the more blue you have after the 12 ms vertical the more you GPU bottlenecked. Same thing with CPU.

Share this post


Link to post
Share on other sites

As for the interpretation of the distribution charts, we should also bear in mind that (even though it is not the case for the charts posted in this thread), the frame time charts do not show values for frame time higher than 30 ms. These are included only in the csv export.

Share this post


Link to post
Share on other sites
2 hours ago, haltux said:

The curve does not represent the history of a frame. The X axis does not represent what happens during a frame. It is a distribution of the frame time.

 

When the blue curve is high, say 6% at 11 ms,  it means that 6% of the frame took 11 ms. So the more blue you have after the 12 ms vertical the more you GPU bottlenecked. Same thing with CPU

 

I didn´t interpreted that as the history of a frame.

It is an histogram. My interpretation was right (as you described).

 

What I was wrong with my interpretation is that I thought that GPU frametimes were including CPU frametimes, and apparently is not like that.

Share this post


Link to post
Share on other sites
3 hours ago, chiliwili69 said:

 

Thanks for sharing this. You are not far from the 90. 

It is good to know that even with the most powerful machines, being at 90 full time is still an utopia. Other panel frequencies like 80Hz (Rift-S, Index) or 72Hz(Pimax) or even 60Hz(Reverb) has better chances to be there.

 

True.  On the other hand I personally can't see or feel any difference above 70Hz.  But I know sensitivity to this varies by person.

Share this post


Link to post
Share on other sites

I can see that even the beefiest PC so far is unable to run the game on 90fps on average. So likely it will be in motion reprojection most of the time (WMR terminology; if the user sets the auto mode which kicks in whenever the fps is between 45 and 90).

 

I wonder about practical visual difference between such a PC and lower specs PC (like mine with i7 4790k) which is in motion reprojection basically all the time.

 

Any thoughts?

 

 

Btw: This benchmark is good also for gauging impact of engine changes on performance on our PCs. Say if the deffered rendering is introduced, we will just re-run the benchmark test and see 👍

Edited by apollon01

Share this post


Link to post
Share on other sites

 

Now my 'core In synch ' and x52 ... ( Tried x53 but not good ) But the test is not much better 

My GPU also OC .... It's not a 2080 ti , but super 

I have read and learned more about bios and oc but I can figure out what more to do , so I guess it's the result my pc 

 

Motherboard:  ROG MAXIMUS CODE
 CPU:                 i5 9600k
 CPU Freq:       5.2 Ghz.    In synch
 L3 cache:        9 MB

 Cores:             6  

 HT:                    Off ( not possible to on/off in Bios ) 
 RAM type:        DDR4
 RAM size:        16 GB ( dual )

 NB Freq:          4300Hz
 RAM Freq:       2x1800 =  3600  MHz (this is 2xDRAM freq if Dual channel)
 RAM Latency:  15  
 GPU:                 2080super
 STMark:          3145

 

Frames: 24922- Time: 180000ms - Avg: 138.456 - Min: 98 - Max: 223

 

Share this post


Link to post
Share on other sites
13 hours ago, apollon01 said:

I wonder about practical visual difference between such a PC and lower specs PC (like mine with i7 4790k) which is in motion reprojection basically all the time

 

I don´t use "motion Smoothness" with my Index. I prefer to lower the settings to be most of the time at 80, and below in heavy scenarios or flying above cities.

I think a top PC will give me significant fps boost in VR to be more time in 80 with a bit better settings. Most likely not always at 80 but for sure a good boost.

 

13 hours ago, apollon01 said:

This benchmark is good also for gauging impact of engine changes on performance on our PCs. Say if the deffered rendering is introduced, we will just re-run the benchmark test and see

 

Frequently new engine version don´t run older flight records. So, this benchmark might became obsolete in the near future. But at least we already got some good info.

13 hours ago, Niiranen-VR said:

CPU Freq:       5.2 Ghz.    In synch

 

Thanks for running again and checking out potential issues.

It still deliver about 18fps less than expected according to greif estimation.

That´s strange.

 

To be sure it is not CPU throtling, you can fix the freq at 5.0GHz or 4.8GHz and run the test again. so we will see if your fps are aligned with the greif expected value.

 

You can also benchmark all your components: CPU, GPU, RAM with publick benchmarks like PAssmark or Userbenchmark (https://www.userbenchmark.com/Software)

Edited by chiliwili69

Share this post


Link to post
Share on other sites
19 hours ago, apollon01 said:

I can see that even the beefiest PC so far is unable to run the game on 90fps on average. So likely it will be in motion reprojection most of the time (WMR terminology; if the user sets the auto mode which kicks in whenever the fps is between 45 and 90).

 

I wonder about practical visual difference between such a PC and lower specs PC (like mine with i7 4790k) which is in motion reprojection basically all the time.

 

I've alternated between trying for a constant 80 FPS and accepting less than that. Especially in a heavy multiplayer situation, I get dips away from 80 that are quite noticeable, and at the moment I have my rig configured to force 40 ASW at all times. I find the consistency is better, but the drawback is trying to ID a plane at 10x that is crossing my field of vision fast, where I will see ghosting. Many of my squadmates are running a GTX 1080 and they also are in 40 ASW most of the time, and they make it work too.

 

With a 4790K you need to accept that reprojection is going to be on most of the time, and tune for that. The nice thing is that 22ms frame time target instead of 11ms is quite a lot more headroom (but you will still see the very occasional frame miss) and you can choose higher settings that look nicer visually. I still wouldn't recommend Ultra or tons of supersample, however, as they might work for you in a low pressure situation and then be too much for your PC to handle in heavy action.

Share this post


Link to post
Share on other sites

lucky   my 2' gen Reverb has a issue with the kabel so back in shop for a while  - kabel/head one point and it goes black - so im not in a hurry to change 'my butttuns' and - no flying 

 

I have here test with my cpu x48 to x52  = 5200 MHz - core in synch  , and one I try not in synch  

There are 2 test - one with fabric gpu and one with fabric OC Gpu ( and the diff is not the top and Avg but the buttom  )

 

and the Last I have tried to manualy OC the Gpu , and It gives the best result 

 

2020-01-20 16:26:09 - Il-2
Frames: 23727 - Time: 180000ms - Avg: 131.817 - Min: 83 - Max: 215    48  in synch 

2020-01-20 16:30:56 - Il-2
Frames: 23374 - Time: 180000ms - Avg: 129.856 - Min: 93 - Max: 213        Gpu oc 

2020-01-20 16:42:48 - Il-2
Frames: 23956 - Time: 180000ms - Avg: 133.089 - Min: 82 - Max: 215    49

2020-01-20 16:46:41 - Il-2
Frames: 24006 - Time: 180000ms - Avg: 133.367 - Min: 95 - Max: 215

2020-01-20 16:56:18 - Il-2
Frames: 24270 - Time: 180000ms - Avg: 134.833 - Min: 83 - Max: 216    50

2020-01-20 17:00:29 - Il-2
Frames: 24367 - Time: 180000ms - Avg: 135.372 - Min: 95 - Max: 218

2020-01-20 17:10:11 - Il-2
Frames: 24633 - Time: 180000ms - Avg: 136.850 - Min: 82 - Max: 216    51

2020-01-20 17:14:21 - Il-2
Frames: 24670 - Time: 180000ms - Avg: 137.056 - Min: 97 - Max: 219

2020-01-20 17:24:48 - Il-2
Frames: 24846 - Time: 180000ms - Avg: 138.033 - Min: 84 - Max: 221    52

2020-01-20 17:28:36 - Il-2
Frames: 24901 - Time: 180000ms - Avg: 138.339 - Min: 99 - Max: 224
-------------------

2020-01-20 17:40:11 - Il-2
Frames: 24654 - Time: 180000ms - Avg: 136.967 - Min: 81 - Max: 217    52 - per core

2020-01-20 17:45:33 - Il-2
Frames: 24167 - Time: 180000ms - Avg: 134.261 - Min: 93 - Max: 217         gpu oc 
-------------------


2020-01-20 23:00:31 - Il-2
Frames: 25165 - Time: 180000ms - Avg: 139.806 - Min: 87 - Max: 231   52 - synch manualy OC 

 

I have tried super position Benchmark and it gives a avg more then the 150 

and tried the userBenchmark :

UserBenchmarks: Game 119%, Desk 110%, Work 88%
CPU: Intel Core i5-9600K - 112.2%
GPU: Nvidia RTX 2080S (Super) - 135%
SSD: Samsung 850 Evo 500GB - 123.4%
SSD: Corsair Neutron GTX 120GB - 65.5%
HDD: Seagate Barracuda 7200.12 1TB - 48.5%
RAM: G.SKILL Trident Z DDR4 3600 C15 2x8GB - 116.8%
MBD: Asus ROG MAXIMUS XI CODE

 

BUT .................. there I see and remember that I have one old harddisk HHD , I dont know wether my game is one this one ...could it be a problem 

Share this post


Link to post
Share on other sites

Thanks for your additional test at different CPU freq.

 

I don´t think the CPU is throttled since you also have the same low performance at 4.8 or 4.9.

 

Your HDD is not the guilty here. A slow hard disk would use more time to load the mission or flight record. But during the run it shouldn´t affect at all since it is loaded into the RAM memory.

 

You userbenchmarks seem to be OK as well. I really have not idea why your PC delivers about 18 fps less than it should be.

Please, double check that you run it exactly with the same settings as indicated in instructions.

Verify that your monitor cable is connected to the GPU, not to the Mobo HDMI/DP output.

Last option would be to do a clean install of IL-2.

 

Could you post pictures of?:

- The interior of the PC itself (just to see Mobo, RAM, CPU cooler, GPU, etc)

- Your current BIOS setup.

- HWMonitor after running the Remagen test, just to see that Min and Max core freq of all cores is the same and not below 5.2GHz

Share this post


Link to post
Share on other sites

Looks like the most recent update has some significant changes to the graphics pipeline, in particular they have changed the texture caching to try to eliminate stutters in multiplayer. Patch 4.004 will probably affect the benchmark results, I'll re-run the benchmarks once I get Combat Box up and working with the new version.

Share this post


Link to post
Share on other sites

Weeeeeel , Il be before Alonzo - I take the challenge up ( that's not there ) 

 

I have checkket again again-  Nvidia g sync - game - and so on and it should be like op 

I have new install steam and game new install to a ssd disk ..... My test is a little bit better , but maked after the new update 

 

I would like to be a average man , but can't figure out too how to raise my social status 🥴🤔

 

I'm on iPad so here's a lot of pic 

0CF84C4D-E514-4A3A-A422-856753EF3FDD.jpeg

B8907FCA-E6CA-4362-8DD4-D1E4044D2A8C.jpeg

01D4159A-ED45-4CED-858F-71BFE43E7A26.jpeg

3F5C6D41-612A-4EC8-A23F-3C3A4EFEAFC2.jpeg

9993B31D-9B38-41C7-B484-93234FA18DDC.jpeg

CF1EBDF6-3D4F-4597-992F-69848DBFC59E.jpeg

Share this post


Link to post
Share on other sites

I did some re-testing. For the 2D version, I don't see any significant change:

DDR4-3600, 15-15-15-36 2T
Game version 4.003  <---- OLD VERSION

Passmark: 3034 / 3029 / 3031

Benchmarks:    Frames / Avg / Min / Max
Run 1:        26895 / 149.417 / 105 / 240
Run 2:        26815 / 148.972 / 105 / 239

DDR4-3600, 15-15-15-36 2T
Game version 4.004  <---- NEW VERSION

Passmark: 3023 / 3033 / 3030

Benchmarks:    Frames / Avg / Min / Max
Run 1:        26865 / 149.250 / 104 / 241
Run 2:        26774 / 148.744 / 101 / 240    

And in fact in the VR benchmark, I also see no real overall improvement, except that my "trying for 80 FPS" test run was able to have fewer frame time spikes. This is only really visible on the graph, not the raw numbers.

 

But actual in-flight performance on Combat Box, on a full server, is way better than previously. I'm able to run at nearly 80 FPS most of the time. I'm probably going to change my strategy from "40 ASW eye candy" to "try for 80, less eye candy".

  • Like 1

Share this post


Link to post
Share on other sites
11 hours ago, Niiranen-VR said:

I'm on iPad so here's a lot of pic 

This all looks good as for the settings go. It is just weird that with those you have core 3 and 4 at 1600 while the others are at 800. They should all be equal.

 

You have the latest BIOS version?

Share this post


Link to post
Share on other sites
On 1/18/2020 at 7:30 PM, chiliwili69 said:
On 1/18/2020 at 5:35 PM, haltux said:

They work in parallel, the CPU works on frame t while the gpu works on frame t-1.

 

Thanks for explaining that. This makes a lot of sense.

Then, it should be possible to reproduce in IL-2 VR these two extreme cases with my Index:

 

CPU bottleneck: (Shadows High, mirror complex) but very low resolution (SS quite below 100%)

CPU frametimes is close to 12.5 (or above) and GPU frametimes below 10 ms

 

GPU bottleneck:(Shadows off, mirror off) but very high resolution (SS quite above 100%)

CPU frametimes is below 10ms and GPU frametimes close to 12.5 (or above)

 

I have to try that with fpsVR

 

I have been testing these two opposite cases:

 

1. Stress CPU, unload GPU: I put SteamVR SS at 50% and did a QMB Kuban in close combat with 5vs8 planes with no clouds. The GPU frametimes was around 7-9 ms and the CPU frametimes was around 9-15 ms.

2. Stress GPU, unload CPU: I punt SteamVRSS at 200%, and did a QMB Lapine 1vs1 with extreme clouds. The frametimes was around 12 to 17 ms and the CPU aorund 5-8 ms

 

So you @apollon01 and @haltux were totally correct. CPU frametimes doesn´t include GPU framtimes, and CPU works in frame n and GPU in frame n-1.

fpsVR is a great tool.

13 hours ago, Niiranen-VR said:

I would like to be a average man , but can't figure out too how to raise my social status 🥴🤔

 

LOL, Thanks for putting those pictures.

 

I think that the reason of your low performance (it is 18 fps lower than expected) is that you run the OC in Auto mode in your BIOS.

Look at "ASUS Multicore Enhacement: Let the BIOS Optimize".

The 800MHz of min freq of HWmonitor confirm that as well.

In Auto mode you leave the BIOS to decide the OC to apply based in CPU load, temps, etc. and it doesn´t guarantee that all your cores are running at 5.2 GHz, sometimess they will do for sure but they will be going up and down along the run of the benchmark track.

 

Most of the people do a fix manual overclock to keep the CPU always at the desired speed independently of the load of the computer. Obviously they have calculated previously the max manual overclock they can achieve in their machines.

 

While you REverb is been replaced, you can enjoy with your soon how to do a moderate manual overclock in your PC.

Try to target first 5.0GHz. I think your i5-9600K with your AIO watercooler can achieve this well.

Overclocking is an iterative process, you need to determine the minimum voltage applied (Vcore voltage in manual) which allows a stable CPU frequency. A stable freq means that you can apply a big CPU load (using a stress tool like Prime95) and you don´t get a Blue-Screen-Of-Death (BSOD).

I don´t know how to do a manual overclock in your ASUS BIOS. Perhaps other people with ASUS can assess you. I am not using BIOS to do OC, I just use MSI Command Center:

 

Share this post


Link to post
Share on other sites
5 hours ago, Alonzo said:

And in fact in the VR benchmark, I also see no real overall improvement, except that my "trying for 80 FPS" test run was able to have fewer frame time spikes. This is only really visible on the graph, not the raw numbers.

 

But actual in-flight performance on Combat Box, on a full server, is way better than previously. I'm able to run at nearly 80 FPS most of the time. I'm probably going to change my strategy from "40 ASW eye candy" to "try for 80, less eye candy".

 

Many thanks Alonzo for running the test in this new game update. We know four things:

- The benchmark still run in 4.004

- It has no impact in fps

- Improves spikes in VR SP

- Improves performance in VR MP

 

That´s really nice.

Share this post


Link to post
Share on other sites

Ceteris paribus guys. CB may have updated their server, less loaded on bandwidth, or simply host a different mission.

 

There is no change to the game in that regard. Unfortunately. We shall be patient.:russian_ru:

Share this post


Link to post
Share on other sites
7 hours ago, SCG_Fenris_Wolf said:

Ceteris paribus guys. CB may have updated their server, less loaded on bandwidth, or simply host a different mission.

 

Well, I run the Combat Box server and we definitely didn't update any of our missions to do performance fixes. We might have had less players or better bandwidth or yes, a different map, but from my perspective (ping 0 to the server) nothing much changed except the game patch, and it improved multiplayer performance a lot.

Share this post


Link to post
Share on other sites
7 minutes ago, Alonzo said:

 

Well, I run the Combat Box server and we definitely didn't update any of our missions to do performance fixes. We might have had less players or better bandwidth or yes, a different map, but from my perspective (ping 0 to the server) nothing much changed except the game patch, and it improved multiplayer performance a lot.

 

Well, you said it yourself: "We might have had less players or better bandwidth or yes, a different map," 

 

Just stating facts, there was no change in the game patch. Maybe tied to a new SteamVR version? 🤔

Share this post


Link to post
Share on other sites
2 hours ago, SCG_Fenris_Wolf said:

 

Well, you said it yourself: "We might have had less players or better bandwidth or yes, a different map," 

 

Just stating facts, there was no change in the game patch. Maybe tied to a new SteamVR version? 🤔

 

I think we are somehow talking at crossed purposes.

 

Game patch 4.004 was released this week. In the Remagen benchmark in both 2D and VR mode, it had no measurable impact on performance. But in multiplayer VR, patch 4.004 has a fix to improve texture management, and this has made a big improvement. Multiplayer VR is the place where this issue arose, and it's where the improvement is felt.

Share this post


Link to post
Share on other sites

Yes I have update BIOS and other stoff in the mashine , new steam and game so no controls or other programs should make a conflict ( its on steam ) - now read more and utupe so this bios is more profound than superficial 

For more than 2 years ago , before I became a gamer, I allso tried to keep my pc fresh and up to date and I had a good feeling when the email was  > wuuush< send quick away ......but now ...

 

The 3 test are same bios 

046D7C3A-255E-4FFE-9ACF-4D33C2FB46D4.jpeg

658A1710-15CC-420F-B065-EB58FE045546.jpeg

FAF06934-1897-4882-B46E-FB0FD6CF8169.jpeg

034D6474-4361-4013-ACDD-5D196F53DE4D.jpeg

8AF1C74E-358F-4B88-B905-92C21329C90F.jpeg

A7453657-D8C2-4FAF-96CA-9869962B38BE.jpeg

FC130C7E-825B-4593-BA46-BE61A9823689.jpeg

F79A61CE-8072-4886-B1B7-A2F46E6997EF.jpeg

B6A10718-F62C-44F3-B192-EE594643C9F7.jpeg

Share this post


Link to post
Share on other sites
On 1/23/2020 at 7:19 AM, Niiranen-VR said:

 

The 3 test are same bios

 

Well done!, I see you improved your performance significantly by applying manual OVerclock at 5.2.

  • Like 1

Share this post


Link to post
Share on other sites

@chiliwili69, as promised - new PC, new tests:

 

Il2 version: 4.004
name: apollon01
Motherboard: ASRock Z390 Taichi Ultimate
CPU: i9 9900K
Active cores: 8
HT: Off
CPU Freq: 5.1 GHz
L3 cache: 16 MB
STMark: 2942
HT: Off
RAM type: DDR4
RAM size: 32 GB
NB Freq: 4300 MHz
RAM Freq: 3600 MHz (Dual channel)
RAM Latency: 16
GPU: 2080Ti
GPU passmark: 16694

 

VR HMD: Reverb
Freq: 90 Hz
SteamVR SS: 102% (2220x2176)

 

Screen
Frames: 28603 - Time: 180000ms - Avg: 158.906 - Min: 73 - Max: 268

 

VR
Frames: 14274 - Time: 180000ms - Avg: 79.300 - Min: 46 - Max: 91

 

I hoped to get over Avg: 160 on screen but I am just a shy off that mark 😐 On the other hand, I am happy to get the i9 stable at 5.1 Ghz @ 1.34 volts. I got water cooling but the PC case is standing in a PC table so the air circulation is probably not optimal. I will call it a day and keep it that way. Performance is pretty solid already!

 

Btw. I am quite surprised to see rather low STMark compared to other processors.

 

Milan

Share this post


Link to post
Share on other sites
On 1/22/2020 at 9:41 AM, chiliwili69 said:

 

Many thanks Alonzo for running the test in this new game update. We know four things:

- The benchmark still run in 4.004

- It has no impact in fps

- Improves spikes in VR SP

- Improves performance in VR MP

 

That´s really nice.

 

I confirm this:  in 4.004 no or negligible FPS impact in the test but significant improvement in VR MP smoothness.

🙂

 

  • Thanks 1

Share this post


Link to post
Share on other sites
6 hours ago, apollon01 said:

I hoped to get over Avg: 160 on screen but I am just a shy off that mark 😐 On the other hand, I am happy to get the i9 stable at 5.1 Ghz @ 1.34 volts. I got water cooling but the PC case is standing in a PC table so the air circulation is probably not optimal. I will call it a day and keep it that way. Performance is pretty solid already!

 

Btw. I am quite surprised to see rather low STMark compared to other processors.

 

Thank you again for running the test in your new PC!

The screen performance is exactly as expected. So don´t worry.

The strange thing is your STMark. It is like your CPU is running at 4.8 GHz. So, please verify that your OC is really maintained constant during the run of the STMArk test.

To check that you can trend the core freqs with MSI afterburner or monitor min/max with HWinfo app.

Air circulation is as important as the CPU cooler, so you should review that as well. You have a top PC, so give it the air it needs.

 

STMark is just very much correlated to CPU freq. That´s why it should be a better variable for the correlation.

  • Thanks 1

Share this post


Link to post
Share on other sites

Make sure you set High Performance in Windows Power Management when running the game or Passmark to insure CPU stays at set clock.

 

Edited by dburne
  • Thanks 1

Share this post


Link to post
Share on other sites

Finally I got set up to run this test and results are below.  I am looking strongly at the possibility of VR and would like to order a HP Reverb (actually I am itching to put in the order today), although my system is not the latest and greatest I'm very interested in feedback as to whether you guys think that's a workable fit for my system:

 

System Specs:

Motherboard:    Gigabyte Z97X-UD5H
CPU:            i7-4790K  
CPU Freq:    4.38 Ghz
L3 cache    8 MBytes
Cores:        4 
HT:        On (8 threads)
RAM Type    DDR3
RAM Size    16 GBytes
NB Frequency    3990.5 Mhz
RAM Freq:    1330 Mhz (2x Dual Channel 665.1)
RAM Latency     10.0
GPU:         1660Ti
STMark:        2630

 

IL2 BoX Release Version: 4.004

 

Results with Graphic Settings as specified:

2020-01-26 11:38:33 - Il-2
Frames: 16687 - Time: 180000ms - Avg: 92.706 - Min: 61 - Max: 175
=================================================
 

Thanks much - Erich aka Stoopy

 

Share this post


Link to post
Share on other sites

S!

 

 EDIT: Moved to own topic. 

Edited by LLv34_Flanker
  • Like 1

Share this post


Link to post
Share on other sites
5 hours ago, =[TIA]=Stoopy said:

although my system is not the latest and greatest I'm very interested in feedback as to whether you guys think that's a workable fit for my system

I was running the same CPU as yours (on turbo 4.5 GHz though) coupled with GTX 1080 and it proved to be all OK for my previous HMD Samsung Odyssey +.

However, the Reverb I got only after I had upgraded the GPU.

 

Quick & dirty comparison of your current GPU and the GTX 1080 shows, that the GTX 1080 is a good deal faster:

https://gpu.userbenchmark.com/Compare/Nvidia-GTX-1660-Ti-vs-Nvidia-GTX-1080/4037vs3603

 

Could you perhaps borrow an HMD from a friend? If so, you could then ramp up SS to Reverb native level and see how your GPU deals with it.

 

My experience with the GTX 1080 on Odyssey + (with SS lower than Reverb's native res) was that the frames dropped only occasionally below 45fps.

Share this post


Link to post
Share on other sites
1 hour ago, apollon01 said:

I was running the same CPU as yours (on turbo 4.5 GHz though) coupled with GTX 1080 and it proved to be all OK for my previous HMD Samsung Odyssey +.

However, the Reverb I got only after I had upgraded the GPU.

 

Quick & dirty comparison of your current GPU and the GTX 1080 shows, that the GTX 1080 is a good deal faster:

https://gpu.userbenchmark.com/Compare/Nvidia-GTX-1660-Ti-vs-Nvidia-GTX-1080/4037vs3603

 

Could you perhaps borrow an HMD from a friend? If so, you could then ramp up SS to Reverb native level and see how your GPU deals with it.

 

My experience with the GTX 1080 on Odyssey + (with SS lower than Reverb's native res) was that the frames dropped only occasionally below 45fps.

 

Thanks.  I'd looked up that comparison before but the 1650 was such a sweet spot in pricing I couldn't justify a 1080.   I pulled the trigger and the Reverb will be here this week so we'll see how it goes.  Have to at least try!

 

  • Like 1

Share this post


Link to post
Share on other sites
15 hours ago, =[TIA]=Stoopy said:

RAM Freq:    1330 Mhz (2x Dual Channel 665.1)

 

Thank you for your test. Your performance is as expected according to Greif correlation.

Independently of the VR device, the guilty guy in your PC is the RAM speed. It is penalizing the whole thing.

You could also OC your 4790K to higher frequencies, for example 4.6 or more depending of your CPU cooling solution.

The Reverb run OK in a 1080Ti, but your GPU is below that, so I am not sure if the GPU would be OK. In any case, if you need GPU power, a GPU is easy to upgrade.

In any case, you can always rely on the "Motion Reprojection" mode of WMR devices. They do a great work on that.

Share this post


Link to post
Share on other sites
On 1/25/2020 at 9:31 PM, chiliwili69 said:

So, please verify that your OC is really maintained constant during the run of the STMArk test.

 

On 1/25/2020 at 10:25 PM, dburne said:

Make sure you set High Performance in Windows Power Management when running the game or Passmark to insure CPU stays at set clock.

Yep, all cores are pegged at 5.1 GHz and the profile is set to High Performance. Still the same result 🤔

 

I also noticed that my NB frequency was lower (4.3 GHz) than the one of other rigs running the i9 9900K. Could that be the culprit? Is that set automatically? Should I increase it? (how & to what frequency?). I have to say I am a bit lost here...

Edited by apollon01

Share this post


Link to post
Share on other sites
10 hours ago, chiliwili69 said:

 

Thank you for your test. Your performance is as expected according to Greif correlation.

Independently of the VR device, the guilty guy in your PC is the RAM speed. It is penalizing the whole thing.

You could also OC your 4790K to higher frequencies, for example 4.6 or more depending of your CPU cooling solution.

The Reverb run OK in a 1080Ti, but your GPU is below that, so I am not sure if the GPU would be OK. In any case, if you need GPU power, a GPU is easy to upgrade.

In any case, you can always rely on the "Motion Reprojection" mode of WMR devices. They do a great work on that.

 

Thanks very much!   I'll look into faster RAM although I believe I am limited to using DDR3.  Will also see how the Reverb works when it arrives later today...gonna be a long week I think. 

 

Share this post


Link to post
Share on other sites
2 hours ago, apollon01 said:

I also noticed that my NB frequency was lower (4.3 GHz) than the one of other rigs running the i9 9900K. Could that be the culprit? Is that set automatically? Should I increase it? (how & to what frequency?). I have to say I am a bit lost here...

 

The NB freq influences the results:

https://forum.il2sturmovik.com/topic/56485-benchmark-for-cpuram-performance-remagen-4002-to-4004/?do=findComment&comment=873373

https://forum.il2sturmovik.com/topic/56485-benchmark-for-cpuram-performance-remagen-4002-to-4004/?do=findComment&comment=873476

 

I am also lost here, but thanks to Alonzo we know more about NB frequency:

https://forum.il2sturmovik.com/topic/56485-benchmark-for-cpuram-performance-remagen-4002-to-4004/?do=findComment&comment=873590

 

My BIOS doesn´t allow me to modify NB freq, but I believe other more modern BIOS you can modify that as radek and Alonzo did.

 

But the STMark result you obtain is a bit strange. At 5.1 you should have around 3090. Look at the "STMArk vs OC" tab in the online sheet.

1 hour ago, =[TIA]=Stoopy said:

I'll look into faster RAM although I believe I am limited to using DDR3.

 

It is difficult to get fast RAM DDR3, but there is still some affordable:

https://pcpartpicker.com/product/jx7wrH/gskill-memory-f32400c11d16gxm

 

Going to that 2400MHz  (check your Mobo can support it) you will gain about 16 fps in monitor.

  • Thanks 2

Share this post


Link to post
Share on other sites
5 hours ago, chiliwili69 said:

It is difficult to get fast RAM DDR3, but there is still some affordable:

https://pcpartpicker.com/product/jx7wrH/gskill-memory-f32400c11d16gxm

 

Going to that 2400MHz  (check your Mobo can support it) you will gain about 16 fps in monitor.

 

Thanks very much for your insight and the parts & pricing info.  Something didn't sound right so I re-checked my build sheet (from 2015!) and saw that I ordered, and indeed have, 2 sticks of Corsair DDR3-2133Mhz RAM.   Looking at the BIOS setting showed it was set to 1330, for whatever odd reason, so a simple setting adjustment has it at 2133.  

 

EDIT:  Re-running the benchmark after putting back in the required GPU and IL2 config settings, I now get these results:

 

2020-01-27 19:51:28 - Il-2
Frames: 18073 - Time: 180000ms - Avg: 100.406 - Min: 68 - Max: 190

 

Not too shabby for a little BIOS correction, and I really do appreciate you catching that, @chiliwili69!

 

 EDIT 2:  My HP Reverb was just dropped off at the door (at 8PM!).  Won't have time to mess with it tonight, and it's frozen solid from being in a cold delivery truck, so the electronics get to come up to room temp and I'll be playing with it tomorrow. Oh boyohboyohboy!!!!

 

 

Edited by =[TIA]=Stoopy

Share this post


Link to post
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

×
×
  • Create New...