Jump to content
chiliwili69

Measuring rig performance: Common Baseline (for IL-2 v3.010)

Recommended Posts

On 5/9/2018 at 5:55 PM, ICDP said:

 

It should be worth noting that WMR devices have their version of reprojection always on, so max FPS will never go much above 45 if a constant 90 FPS cannot be maintained.  Hence the max FPS being comparatively low using WMR devices and the average being around 45 FPS.  The main thing to take away from this result is that I was stuck at avg 45 FPS for most of the track and it always felt smooth due to the reprojection feature.  It is also worth noting that the results for SS set to 130% and 170% are within margin of error on the min FPS.

 

 

Why would WMR set FPS to 45 if the refresh rate of the Odyssey is 60 or 90? 

Share this post


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

 

Why would WMR set FPS to 45 if the refresh rate of the Odyssey is 60 or 90? 

 

I think ASW automatically kicks in below around 60 FPS which limits FPS at 45.  If I lower settings it will sit from about 60-90 FPS.  It behaves very different to the CV1, I think it's because it can work at 60, or 90Hz.

Share this post


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

 

I think ASW automatically kicks in below around 60 FPS which limits FPS at 45.  If I lower settings it will sit from about 60-90 FPS.  It behaves very different to the CV1, I think it's because it can work at 60, or 90Hz.

 

I think ASW for Oculus limits to 45fps because of 45/90 fps native refresh rate of Oculus. It would be weird if it tried to force Odyssey to 45fps as that would create all kinds of stutter because unlike Oculus, Odyssey has a 60/90fps refresh rate. 

Share this post


Link to post
Share on other sites

That's not how the Odyssey, or other WMR devices work.  WMR devices will work at 90Hz if you have HDMI 2.0, or 60Hz with older HDMI ports.

 

So the WMR version of ASW will be either 30/60 or 45/90 depending upon the HDMI standard you have.

Edited by ICDP

Share this post


Link to post
Share on other sites
On 5/30/2018 at 5:22 PM, ICDP said:

That's not how the Odyssey, or other WMR devices work.  WMR devices will work at 90Hz if you have HDMI 2.0, or 60Hz with older HDMI ports.

 

So the WMR version of ASW will be either 30/60 or 45/90 depending upon the HDMI standard you have.

 

 

Are you 100% sure? Setting ASW to 45 on a device with a 6ohz refresh rate would surely cause problems. 

Edited by ClarkFable

Share this post


Link to post
Share on other sites

Chiliwili69, thanks so much for coordinating this effort! I think my results are a little embarrassing, but maybe you and the gang can clue me if I may be doing something wrong. Or let me know if, as I suspect, my GTX 1060 is what's holding me back. I noticed that Coconut's results with the same CPU (i5-4690K oc'd at 4.5GHz) and a 980ti are significantly better.

 

Here are my results with IL2 version 3.003, using my HTC Vive:

 

Frames 6304, Avg 31.52, Min 23, Max 45

Public Passmark 7644/2237 (stock is 3.5 GHz)

User Passmark  10529/2698 (oc'd to 4.5 GHz)

 

I look forward to your feedback!

 

 

Share this post


Link to post
Share on other sites

I got the Odyssey. It runs at 90Hz, and at 45Hz with ASW. WMRs use ASW, the same tech Oculus Rift uses, they don't have "their own reprojection feature". If you turn down the WMR to 60Hz, its FOV gets bugged as before the fix in 3.003 by the developers. ASW there halves the framerate to 30Hz.

Share this post


Link to post
Share on other sites
On 5/31/2018 at 10:48 PM, ClarkFable said:

 

 

Are you 100% sure? Setting ASW to 45 on a device with a 6ohz refresh rate would surely cause problems. 

 

It would be 60/30 or 90/45 depending upon HDMI port type.

Edited by ICDP
Just saw Fenris answered this above.

Share this post


Link to post
Share on other sites

Precisely as ICDP said.

 

You can also manually enable 60Hz. But 60Hz is not an option, because it messes up the field of view/scale of things - fish-eye view, and everything is tiny.

Share this post


Link to post
Share on other sites
23 hours ago, Jack59 said:

I look forward to your feedback!

thanks Jack59 for running your test.

Don´t be embarrassed, all test data is very valuable, the low fps and the high fps. The important thing is that we can learn what it is good for IL-2 VR and what is not.

 

There are two factor for your low performance:

 

1. This is my fault. I have just realized now that the Vive is using a different internal supersampling ratio than the Rift. So even if they have the same pixels in the physical display, when you put 170% in the Rift, it is not the same than putting 170% on the Vive. So, the test with Vive has to be run at 144% SS in SteamVR (or 1.2 in OTT) to have the same number of pixels in the render. So, if you can repeat your test with 144% you will achieve a bit better result.

https://forum.il2sturmovik.com/topic/36160-supersampling-ratios-rift-vs-vivepro-vs-odyssey/

 

2. The 1060 card performs lower than the 980Ti. In fact, the 1060 card is botlenecking this test a lot. You can try to run the same test at lower SS (100% or 70% or lower) until you will see that your Avg fps is not increased anymore. Then, this will be the fps that your current PC (CPU&RAM) could deliver if your GPU would not be constraining the test. So, you can now the fps that your system will deliver if you would have a 1080Ti for example.

 

Being said that, it doesn´t mean that your system is not fully valid for VR. IF you run the test at SS=100% and with Balanced or Low settings you will see that your performance could be valid. On top of that, the test is just a benchmark test, so we put many planes, smoke and fire. This is not the normal playing.

On 5/9/2018 at 11:55 PM, ICDP said:

Supersampling set to 132% to match 7.3 mega pixels equivalent to CV1 tests at 170% SS.

Frames: 8935 - Time: 200000ms - Avg: 44.675 - Min: 40 - Max: 52

 

According to this table you should use 144% to have 7.3 mega pixels:

https://forum.il2sturmovik.com/topic/36160-supersampling-ratios-rift-vs-vivepro-vs-odyssey/

 

Right?

  • Thanks 1

Share this post


Link to post
Share on other sites
37 minutes ago, chiliwili69 said:

thanks Jack59 for running your test.

Don´t be embarrassed, all test data is very valuable, the low fps and the high fps. The important thing is that we can learn what it is good for IL-2 VR and what is not.

 

There are two factor for your low performance:

 

1. This is my fault. I have just realized now that the Vive is using a different internal supersampling ratio than the Rift. So even if they have the same pixels in the physical display, when you put 170% in the Rift, it is not the same than putting 170% on the Vive. So, the test with Vive has to be run at 144% SS in SteamVR (or 1.2 in OTT) to have the same number of pixels in the render. So, if you can repeat your test with 144% you will achieve a bit better result.

https://forum.il2sturmovik.com/topic/36160-supersampling-ratios-rift-vs-vivepro-vs-odyssey/

 

2. The 1060 card performs lower than the 980Ti. In fact, the 1060 card is botlenecking this test a lot. You can try to run the same test at lower SS (100% or 70% or lower) until you will see that your Avg fps is not increased anymore. Then, this will be the fps that your current PC (CPU&RAM) could deliver if your GPU would not be constraining the test. So, you can now the fps that your system will deliver if you would have a 1080Ti for example.

 

Being said that, it doesn´t mean that your system is not fully valid for VR. IF you run the test at SS=100% and with Balanced or Low settings you will see that your performance could be valid. On top of that, the test is just a benchmark test, so we put many planes, smoke and fire. This is not the normal playing.

 

According to this table you should use 144% to have 7.3 mega pixels:

https://forum.il2sturmovik.com/topic/36160-supersampling-ratios-rift-vs-vivepro-vs-odyssey/

 

Right?

 

Wow, great info......thanks!

 

I'll be sure to rerun the test at 144% within the next 24 hours, and post the results. Relative to your comment about my system still being valid for VR, yes, and I had run your test with my (lowered) settings (with included 1.0 SS) and got the following results: Frames 12435, avg 62.175, min 43, max 90. As you can well imagine, the image quality leaves much to be desired. But, VR is so absolutely, unequivocally awesome, that I'll still take it over far superior monitor graphics.

 

One quick question: Going forward, assuming that I keep my same cpu but wish to improve my gpu, is it a question of matching a gpu to my cpu capabilities, with there being such a thing as gpu overkill? Or is it that no matter how good of a card I get, I will enjoy greater and greater capability in say for example, SS? In other words, is it possible that a 1080ti will be overkill as opposed to a 1070 or 1080?

Share this post


Link to post
Share on other sites
10 hours ago, Jack59 said:

One quick question: Going forward, assuming that I keep my same cpu but wish to improve my gpu, is it a question of matching a gpu to my cpu capabilities, with there being such a thing as gpu overkill? Or is it that no matter how good of a card I get, I will enjoy greater and greater capability in say for example, SS? In other words, is it possible that a 1080ti will be overkill as opposed to a 1070 or 1080?

 

Going to a higher GPU will help you to put a higher SS number, which will give you a bit more detailed graphics. But it might not help in allowing you to go to HIGH or ULTRA graphics settings since it is dependent on CPU performance.

 

With a 1080 or 1080Ti, you will pass the performance ball to the roof of the CPU. 

Once the hardware is fixed (ie, for a given GPU and CPU/RAM system), then the settings has to be tweaked to load GPU and CPU equally. Your manipulated variables will be SS (for loading GPU) and IL-2 settings (for loading CPU).

 

In your case I would go for a 1080 rather than a 1080Ti. But it is also true that for the cost of a 1080 you can buy a top Mobo+CPU+RAM. But then your system will be fully bottlenecked by the 1060.

  • Thanks 1

Share this post


Link to post
Share on other sites
13 hours ago, Jack59 said:

 

One quick question: Going forward, assuming that I keep my same cpu but wish to improve my gpu, is it a question of matching a gpu to my cpu capabilities, with there being such a thing as gpu overkill? Or is it that no matter how good of a card I get, I will enjoy greater and greater capability in say for example, SS? In other words, is it possible that a 1080ti will be overkill as opposed to a 1070 or 1080?

 

IMHO no, with your rig a 1080 Ti would not be overkill.

 

If you have the funds and that is not an issue, I would suggest going with the 1080 Ti. VR is very demanding and will take all the GPU horsepower one can throw at it.

I went from a 1080 to a 1080 Ti  last year and even with that saw a  boost in my performance.  I certainly did not regret getting it and was glad I did.

 

Understand even a 1080 Ti is not enough to give a constant 90 fps in VR with all settings maxed out, systems just do not exist today that will. Combat flight sims are very demanding, especially in VR.

 

  • Thanks 1

Share this post


Link to post
Share on other sites

chiliwili69 and dburne, thank you both for your valuable inputs! I have some decisions to make. But I do want to move forward with my hardware.

 

I reran the test with 144% SS for my Vive. The average is less scary: Frames 8195 Avg 40.975, Min 25, Max 46

 

chiliwili69, as you suggested I kept on repeating the test, progressively reducing the SS until I stopped getting an increase in avg fps. I continued to enjoy higher fps all the way down to the minimum setting of 20% SS, where my average was 58.86. If I could reduce the SS further maybe the average frames would go a bit higher, maybe not. But I guess the testing has already shown what a bottleneck the 1060 truly is!

Share this post


Link to post
Share on other sites
4 hours ago, Jack59 said:

I continued to enjoy higher fps all the way down to the minimum setting of 20% SS, where my average was 58.86. If I could reduce the SS further maybe the average frames would go a bit higher, maybe not. But I guess the testing has already shown what a bottleneck the 1060 truly is!

 

58 fps is actually what your fps should be based just in your CPU single-thread performance (fps expected column).

 

There is a similar case from BOO who went from 1060 to 1080 in monitor:

https://forum.il2sturmovik.com/topic/36882-advice-please-msi-1080-gaming-about-to-buy/

 

Whatever you go, 1080 or 1080Ti you will benefit well from it.

In my case I went from 1070 to 1080Ti and I got zero gain in the IL-2 VR test and test settings. It doesn´t mean you I would have zero gain with other settings.

 

  • Thanks 1

Share this post


Link to post
Share on other sites

OK Gentlemen, with your much appreciated recommendations stirred around in the pot, I've finally made a decision. I've decided to go with a 1080ti, and now have it on order. I have no doubt that the 1080 would have been a decent match with my current cpu, and very nearly went that route. But I'm hoping that the 1080ti will be a good match with any cpu that I may upgrade to in the near future. At least that's the plan...

 

When I get my 1080ti all settled in I'll be sure to rerun the test and post my results.

 

 

Share this post


Link to post
Share on other sites
8 minutes ago, Jack59 said:

OK Gentlemen, with your much appreciated recommendations stirred around in the pot, I've finally made a decision. I've decided to go with a 1080ti, and now have it on order. I have no doubt that the 1080 would have been a decent match with my current cpu, and very nearly went that route. But I'm hoping that the 1080ti will be a good match with any cpu that I may upgrade to in the near future. At least that's the plan...

 

When I get my 1080ti all settled in I'll be sure to rerun the test and post my results.

 

 

 

:good:

 

Good choice, there really is no GPU currently better for VR.

And I doubt the initial offering of the 1100 series cards will be that much better than a 1080 Ti, probably not until a Ti version which usually is around a year later.

 

Just remember it is not how it performs in a benchmark test, but how it performs whilst you actually play the game that counts. They can be vastly different as I discovered.

Edited by dburne

Share this post


Link to post
Share on other sites
8 hours ago, Jack59 said:

But I'm hoping that the 1080ti will be a good match with any cpu that I may upgrade to in the near future.

 

If budget allows, good choice. With that GPU you will not be GPU constrained with any of the current VR devices and will allows higher levels of SS.

I believe you should achieve a value closer to 58 fps avg in the test. Or a bit higher, who knows...

Share this post


Link to post
Share on other sites
On 6/3/2018 at 8:18 PM, chiliwili69 said:

According to this table you should use 144% to have 7.3 mega pixels:

https://forum.il2sturmovik.com/topic/36160-supersampling-ratios-rift-vs-vivepro-vs-odyssey/

 

Right?

 

Right below that quote I posted almost identical results for 170% supersampling.

 

Supersampling set to 132% to match 7.3 mega pixels equivalent to CV1 tests at 170% SS.

Frames: 8935 - Time: 200000ms - Avg: 44.675 - Min: 40 - Max: 52

 

Supersampling set to 170%

2018-05-09 22:36:37 - Il-2
Frames: 8954 - Time: 200000ms - Avg: 44.770 - Min: 39 - Max: 58

Share this post


Link to post
Share on other sites

A little late to the party but here it goes :

IL-2 v3.003

Oculus Rift

8700K@5.0Ghz

1080Ti Founders Edition

ASUS ROG STRIX Z370-F GAMING

32Gb DDR4@3300Mhz

CPU Mark : 16899

Single thread : 3034

Frames: 14064 - Time: 200000ms - Avg: 70.320 - Min: 42 - Max: 91

 

Note : Hyperthreading had negative effect, min frames dropped to 40.

            1080Ti OC profile had zero impact in performance.

 

Edited by Horna

Share this post


Link to post
Share on other sites

I've completed the installation of my new GTX 1081ti (replacing my GTX 1060). Wow! A significant improvement in fps and in my ability to increase my graphics settings.

 

My new numbers: Frames 11137, Avg 55.685, Min 42, Max 90.

 

Chili, your avg fps estimate was close!

 

I am now cpu bottlenecked, with my 4690K holding me back. But that doesn't change the fact that I now have a much improved, and to me very satisfactory setup. Again my thanks to you gents for your expertise and for encouraging me to go this route....

Share this post


Link to post
Share on other sites

A bit of an update : Did some tests and it seems my results were hampered by VRM throttling, figured it out when going straight to 5.1Ghz and my avg FPS dropped to 65. I have now a solid 5.1 OC so feel free to update my result. The minimum FPS is quite a killer :(

 

2018-06-09 10:18:42 - Il-2
Frames: 14816 - Time: 200000ms - Avg: 74.080 - Min: 43 - Max: 91

 

CPU MARK : 17291

SINGLE THREAD : 3102

 

Share this post


Link to post
Share on other sites
On 6/6/2018 at 9:34 PM, ICDP said:

Right below that quote I posted almost identical results for 170% supersampling.

 

Supersampling set to 132% to match 7.3 mega pixels equivalent to CV1 tests at 170% SS.

Frames: 8935 - Time: 200000ms - Avg: 44.675 - Min: 40 - Max: 52

 

Supersampling set to 170%

2018-05-09 22:36:37 - Il-2
Frames: 8954 - Time: 200000ms - Avg: 44.770 - Min: 39 - Max: 58

Thanks for repeating the test.

That´s mean that your GPU is far from being the bottleneck.

On 6/9/2018 at 1:18 AM, Jack59 said:

My new numbers: Frames 11137, Avg 55.685, Min 42, Max 90.

That´s great improvement!

Thanks to you indeed for demonstrating how a 1060 can affect the performance.

On 6/9/2018 at 9:40 AM, Horna said:

Did some tests and it seems my results were hampered by VRM throttling, figured it out when going straight to 5.1Ghz and my avg FPS dropped to 65. I have now a solid 5.1 OC so feel free to update my result.

Thanks for reporting your test with 5.0 and 5.1. You got about 4 fps gain in this complex scenario. That´s a good "free" gain if your temps are OK.

You are now at the top performer! :cool:

 

How did you know that the Voltage Regulator was throtling the CPU? and how did you solve that?

 

PD: I have been doing also some test at 4.9 and 5.0 and I get some negative gains with respect to running at 4.8. So, I might have the same problem with the Voltage Regulator Module.

 

  • Upvote 1

Share this post


Link to post
Share on other sites

Hi all,

I have included my results below.

My motivation is that having experienced Il-2 in VR, there is no going back; yet I am disappointed in the performance I am getting with my rig.

Now that I have done some reading of the forum, I understand that I am most likely bottlenecked by my poor single thread performance.

I find that single player career mode is nearly unplayable due to fps (or lack of), and even in multiplayer, when other aircraft are close (<1km), there is a noticeable amount of blurring/ghosting of the aircraft.

I am considering upgrading to an 8700k for the sole purpose of getting better performance in VR in Il-2 (and I may start playing DCS as well).

However, given that it is a sizeable chunk of money for me to spend (AUD$1000), I want to be sure that I'll get noticeable performance improvements.

What do you recommend? Would it be worth the upgrade? Or am I better of spending my money on other hardware upgrades...like a sweet set of MFG Crosswinds.

Any feedback is appreciated.

Cheers.

 

IL-2: v3.003

CPU: Ryzen 7 1700 OC @ 3.9 GHz

mobo: Gigabyte AB350 Gaming 3

RAM: 16 GB @ 3193.2 MHz

GPU: GTX 1080

Oculus Rift CV1

 

FRAPS results: 

Frames: 8842 - Time: 200000ms - Avg: 44.210 - Min: 32 - Max: 68

 

Passmark Results:

CPU Mark: 10314

CPU Single Threaded: 2076

Edited by KIWIvolshebnik
forgot mobo

Share this post


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

Thanks for repeating the test.

That´s mean that your GPU is far from being the bottleneck.

That´s great improvement!

Thanks to you indeed for demonstrating how a 1060 can affect the performance.

Thanks for reporting your test with 5.0 and 5.1. You got about 4 fps gain in this complex scenario. That´s a good "free" gain if your temps are OK.

You are now at the top performer! :cool:

 

How did you know that the Voltage Regulator was throtling the CPU? and how did you solve that?

 

PD: I have been doing also some test at 4.9 and 5.0 and I get some negative gains with respect to running at 4.8. So, I might have the same problem with the Voltage Regulator Module.

 

 

I got first  a bit suspicious when looking at the chart, I got highest Passmark results but lowest FPS compared to the top 5.

So I tried the 5.1ghz OC and while Passmark results got better, the avg FPS in the test dropped even more, to 66 FPS!

As AIDA64 benchmark showed no CPU throttling, there was not much left to suspect but the VRM.

 

These are the settings I changed in BIOS, but do mind that I'm not by any means an expert overclocker so I can't say if these settings are safe or not. Basically I've been roaming through zillion websites and youtube channels searching for info for like, 2 weeks or so 😓

CPU LOAD LINE CALIBRATION LVL 6

CPU CURRENT CAPABILITY 100% => 130%

VRM SPREAD SPECTRUM => DISABLED

CPU POWER DUTY CONTROL : STANDARD => EXTREME

CPU POWER PHASE CONTROL : STANDARD => EXTREME

Last, I installed a 50mm fan blowing air to VRM module, I read that it helps greatly and is also recommended by the mobo manufacturer (ASUS)

 

P.S. That chart is awesome work, without it I'd been happily flying with worse performance while just wasting energy! Kudos to the other testers too, the info is invaluable!

 

 

 

 

  • Upvote 1

Share this post


Link to post
Share on other sites
18 hours ago, KIWIvolshebnik said:

Hi all,

I have included my results below.

My motivation is that having experienced Il-2 in VR, there is no going back; yet I am disappointed in the performance I am getting with my rig.

Now that I have done some reading of the forum, I understand that I am most likely bottlenecked by my poor single thread performance.

I find that single player career mode is nearly unplayable due to fps (or lack of), and even in multiplayer, when other aircraft are close (<1km), there is a noticeable amount of blurring/ghosting of the aircraft.

I am considering upgrading to an 8700k for the sole purpose of getting better performance in VR in Il-2 (and I may start playing DCS as well).

However, given that it is a sizeable chunk of money for me to spend (AUD$1000), I want to be sure that I'll get noticeable performance improvements.

What do you recommend? Would it be worth the upgrade? Or am I better of spending my money on other hardware upgrades...like a sweet set of MFG Crosswinds.

Any feedback is appreciated.

Cheers.

 

IL-2: v3.003

CPU: Ryzen 7 1700 OC @ 3.9 GHz

mobo: Gigabyte AB350 Gaming 3

RAM: 16 GB @ 3193.2 MHz

GPU: GTX 1080

Oculus Rift CV1

 

FRAPS results: 

Frames: 8842 - Time: 200000ms - Avg: 44.210 - Min: 32 - Max: 68

 

Passmark Results:

CPU Mark: 10314

CPU Single Threaded: 2076

 

After upgrading from 4.5ghz overclocked 4770K to 5.1 Ghz overclocked 8700K I can say that the performance of the new rig is excellent compared to the old one.

No more stutters when switching technochat on/off, fluent flying near ground compared to the 40 FPS struggle with the old one. Basically the upgrade changed my flying from continuous settings tinkering and frustration to an enjoyable simming experience.

That said, I think I wasted money on the 8700K, I'd buy a 8600K now, after done some research. For me, the hyperthreading of the 8700K works only as a heat generator, it will heat up the CPU by 10+ degrees celsius minimum while doing nothing to improve game performance. Tested it on DCS, IL-2 and Total War : Warhammer 2. Zero gain.

So, if you are going to bite the bullet and go for the upgrade, consider the 8600K also. From Silicon Lottery site, those have good OC capability too :

 

"This CPU is guaranteed stable when using the settings below and matching components from our QVL. We go through a rigorous stress test routine to ensure stability for the vast majority of use cases. 

CPU Multiplier: 50

BCLK: 100.0

CPU Vcore: 1.400V 

AVX Offset: 2

As of 6/08/18, the top 88% of tested 8600Ks were able to hit 5.0GHz or greater.

This CPU includes our delidding service! Factory thermal paste underneath the IHS is replaced with Thermal Grizzly Conductonaut. Depending on the workload, this typically lowers peak core temperatures anywhere from 15°C to 25°C."

 

Share this post


Link to post
Share on other sites

You can easily disable the hyperthreading in the bios. It has been that way ( heat generator with no real benefit for gaming) for some time.

I have had it disabled on my i7 4820K since shortly after doing the build.

 

If I were buying today I would go for the 8700K in a heartbeat. But that's just me.

Trying to hold out till latter part of the year though.

 

 

Edited by dburne

Share this post


Link to post
Share on other sites
3 minutes ago, dburne said:

You can easily disable the hyperthreading in the bios. It has been that way ( heat generator with no real benefit for gaming) for some time.

I have had it disabled on my i7 4820K since shortly after doing the build.

 

If I were buying today I would go for the 8700K in a heartbeat. But that's just me.

Trying to hold out till latter part of the year though.

 

 

Yeah I have it disabled. I bash myself because I paid for it 😄

Share this post


Link to post
Share on other sites
2 minutes ago, Horna said:

Yeah I have it disabled. I bash myself because I paid for it 😄

 

That's ok, pretty sure that probably is not the only difference so don't bash yourself too hard.

  • Upvote 1

Share this post


Link to post
Share on other sites
20 hours ago, KIWIvolshebnik said:

What do you recommend? Would it be worth the upgrade?

 

A 8600K with good cooling will greatly improve your experience and will save you some money with respect to the 8700K.

You can sell you CPU+Mobo at a good price, so the net upgrade should less than 1000AUD.

Otherwise you can try to play with Low or Balanced settings and buy that nice pedals.

 

Share this post


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

After upgrading from 4.5ghz overclocked 4770K to 5.1 Ghz overclocked 8700K I can say that the performance of the new rig is excellent compared to the old one.

No more stutters when switching technochat on/off, fluent flying near ground compared to the 40 FPS struggle with the old one. Basically the upgrade changed my flying from continuous settings tinkering and frustration to an enjoyable simming experience.

That has pretty much convinced me. I am continuously adjusting SS and graphics settings, trying to find an acceptable compromise between quality and performance.

 

7 hours ago, Horna said:

That said, I think I wasted money on the 8700K, I'd buy a 8600K now, after done some research. For me, the hyperthreading of the 8700K works only as a heat generator, it will heat up the CPU by 10+ degrees celsius minimum while doing nothing to improve game performance. Tested it on DCS, IL-2 and Total War : Warhammer 2. Zero gain.

So, if you are going to bite the bullet and go for the upgrade, consider the 8600K also.

5

I hadn't considered the 8600K, but I'll definitely have a look now. Cheers.

 

5 hours ago, chiliwili69 said:

A 8600K with good cooling will greatly improve your experience and will save you some money with respect to the 8700K.

You can sell you CPU+Mobo at a good price, so the net upgrade should less than 1000AUD.

Otherwise you can try to play with Low or Balanced settings and buy that nice pedals.

Thanks. I'll definitely look at the 8600K. The money saved would start me on the way to the pedals as well.

I've tried with Low/Balanced settings, but I need the pretty visuals. :)

Share this post


Link to post
Share on other sites

So after upgrading, here are my new results.

 

IL-2: v3.004

CPU: 8600K OC @ 5.0 GHz

mobo: ASRock Z370 Taichi

RAM: 16 GB @ 3199.2 MHz (interestingly, the same RAM at the same XMP Profile, but running ever so slightly higher)

GPU: GTX 1080

Oculus Rift CV1

 

FRAPS results: 

Frames: 13801 - Time: 200000ms - Avg: 69.005 - Min: 42 - Max: 91

 

Passmark Results:

CPU Mark: 16025

CPU Single Threaded: 3019

 

Overall, I am very happy with the improvement. Haven't played much yet, but looking forward to TAW.

In the little that I have played, I've found that fighters are pretty seamless, but the German bombers are still down in the mid 40 FPS.

 

A big thank you to those that gave me advice and feedback, it is much appreciated.

Share this post


Link to post
Share on other sites
On 6/25/2018 at 7:40 AM, KIWIvolshebnik said:

Avg: 69.005 - Min: 42 - Max: 91

 

Congratulations for your new achievement!

 

Yes!, the i5-8600K (6 cores) is a very good choice for IL-2. It can be overclocked well and your performance is as good as the flaship CPU i7-8700K (6 cores + 6 logical cores), and you saved a nice money!

 

Thank you to you also for reporting your nice results. This is the nice thing of the benchmarks, your data can be seen by many other people who are thinking to upgrade their PCs.

 

If you have some time during this summer, you could try to run the test at different overclocking (3.6, 3.8, 4.0,... until 5.0), just to know how worth is every step in OC.

  • Thanks 1

Share this post


Link to post
Share on other sites

For real world gaming experience I myself would still lean toward the 8700k , or maybe even the newer 8086k. For VR  I want the best I can get.

I will be doing a new build myself later this year, still not sure of exact timing - I have the funds now but am holding off a bit and see what might be coming down the pike.

 

Edited by dburne

Share this post


Link to post
Share on other sites
20 minutes ago, dburne said:

For real world gaming experience I myself would still lean toward the 8700k , or maybe even the newer 8086k. For VR  I want the best I can get.

 

For IL-2 VR the OC frequency is the king (assuming the CPU has 4 cores or more and you have 1070 GPU or above). So, i5-8600K or i7-8700K or i7-8086K at the same freq. will produce same performance in IL-2 VR.

 

For other games I really don´t know.

 

If you are going to have a new CPU you can take a look of the delided (+warranty) CPUs for cofee lake (including 8086K):

https://siliconlottery.com/collections/coffeelake

 

There is 450$ difference between the i5-8600K at 5.3GHz (but sold out) and the i7-8086K at 5.3GHz.

Share this post


Link to post
Share on other sites
On 6/25/2018 at 8:40 AM, KIWIvolshebnik said:

So after upgrading, here are my new results.

 

IL-2: v3.004

CPU: 8600K OC @ 5.0 GHz

mobo: ASRock Z370 Taichi

RAM: 16 GB @ 3199.2 MHz (interestingly, the same RAM at the same XMP Profile, but running ever so slightly higher)

GPU: GTX 1080

Oculus Rift CV1

 

FRAPS results: 

Frames: 13801 - Time: 200000ms - Avg: 69.005 - Min: 42 - Max: 91

 

Passmark Results:

CPU Mark: 16025

CPU Single Threaded: 3019

 

Overall, I am very happy with the improvement. Haven't played much yet, but looking forward to TAW.

In the little that I have played, I've found that fighters are pretty seamless, but the German bombers are still down in the mid 40 FPS.

 

A big thank you to those that gave me advice and feedback, it is much appreciated.

Gratz for the new rig! And as suspected, the 8600K won't fall behind when compared to 8700K. Good to get the 8600K on the chart too o7

  • Thanks 1

Share this post


Link to post
Share on other sites
22 minutes ago, chiliwili69 said:

 

For IL-2 VR the OC frequency is the king (assuming the CPU has 4 cores or more and you have 1070 GPU or above). So, i5-8600K or i7-8700K or i7-8086K at the same freq. will produce same performance in IL-2 VR.

 

For other games I really don´t know.

 

If you are going to have a new CPU you can take a look of the delided (+warranty) CPUs for cofee lake (including 8086K):

https://siliconlottery.com/collections/coffeelake

 

There is 450$ difference between the i5-8600K at 5.3GHz (but sold out) and the i7-8086K at 5.3GHz.

 

For IL-2 VR running specifically this benchmark program the OC frequency is the king (assuming the CPU has 4 cores or more and you have 1070 GPU or above).

 

There corrected that for ya.

;)

I myself don't play benchmarks...

 

Yeah I have been looking hard at that 8086k , was looking into the delidding tool it does not seem to be overly complicated process, nice to know where I can order one already de-lidded though. That might suit me better. Tempting tempting.

 

Edited by dburne

Share this post


Link to post
Share on other sites
19 minutes ago, dburne said:

 

For IL-2 VR running specifically this benchmark program the OC frequency is the king (assuming the CPU has 4 cores or more and you have 1070 GPU or above).

 

There corrected that for ya.

In what situation is CPU frequency not the most important?

 

19 minutes ago, dburne said:

I myself don't play benchmarks...

How do you analyze performance?

Edited by Mitthrawnuruodo

Share this post


Link to post
Share on other sites
14 minutes ago, Mitthrawnuruodo said:

In what situation is CPU frequency not the most important?

 

How do you analyze performance?

 

There is more at play than just raw cpu frequency. Granted it has a large effect. Memory, memory speed, GPU memory, GPU clock and memory speed,cache size,  etc.

 

I analyze performance by how it performs for me actually running the game. There is a difference between actually running the game , looking and flying around, versus watching a trk file play. 

If I based my decisions off of this benchmark I would have sold the Rift and gone back to monitor.

Dang glad I did not.

Edited by dburne

Share this post


Link to post
Share on other sites
11 hours ago, dburne said:

I myself don't play benchmarks...

 

:dash:

 

Dburne, my friend, you are tougher than me. 

I am from a region (Aragon) where the people is known by being very tough minded, but you really beat me with that recurrent topic.

 

When you run a Flight record, the IL-2 game engine do exactly the same calculations in CPU and GPU that you do when doing normal play game.

If you don´t believe that you can do a simple test:

 

1.- Select a quick mission and press Autopilot having FRAPS running in the the background. Then annotate the avg fps or the fps of every second of the mission

2.- Repeat the same quick mission and press Recording and Autopilot.

3.- Load the flight track of point two having FRAPS in the background. Then compare the results with point 1. You will have exactly the same results.

 

Now, another matter is if the benchmark used in this test represent the avg fps in a normal mission play. The answer is NO. The benchmark track contains a lot of planes and smoke and bullets, it is just 3 minutes of very heavy combat.

 

Edited by chiliwili69

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...