Jump to content
chiliwili69

Benchmark for CPU/RAM performance: Remagen 4.002 to 4.005

Recommended Posts

2 минуты назад, chiliwili69 сказал:

Your preformance is quite beyond expected. Probably your RAM is helping.

Yes, but online it helps a little. Like everyone else, constant stuttering and fps drops to 60-65. Solid 80 (index) is very small. Online stuttering is almost independent of the settings.High or medium makes no difference.

Share this post


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

I will try to do some more testing

 

WOW! what a nice collection of rig. Al  of them would be interesting, specially the Rig1 and Rig2. You can try to disable some cores of the CPU to low temp and reach higher freq. IL-2 doesn´t need more than 4 cores.

9 minutes ago, WallterScott said:

Probably your RAM is helping

 

It would be interesting to see how your PC performs at other RAM frequencies 3200/3600/4000 for example, if they are available in your XMP profiles.

Share this post


Link to post
Share on other sites

Motherboard: ASRock Z270 Pro4
 CPU:                 7600K
 CPU Freq:        5 Ghz
 L3 cache:        6 MB
 RAM type:        DDR4
 RAM size:        16 GB
 RAM Freq:        2933 MHz
 RAM Latency: 16
 GPU:                 1080Ti
 STMark:           2940

 

Frames: 24393 - Time: 180000ms - Avg: 135.517 - Min: 94 - Max: 217

Share this post


Link to post
Share on other sites

I have just run a couple of experiments with Grass and Shadows.

 

GRASS: There is very very little effect on this track since we have not landed/takeoff. In the low flight before crossing the bridge we can see the the Ultra delivers about 5 fps less, but the rest of the track is almost identical. So for flights above 300-400 meters the Grass has not influence.

Grass.thumb.png.cbfe91a38060d3dccc99944b03569172.png

 

SHADOWS: Here the effect is more pronounced as there is a lot of shadows in the cockpit and ground.

First at all, the three first level (OFF, LOW, MEDIUM) produce almost identical results. Then when going to High we lose 10fps on average. And another 6 fps on average when going to Ultra

1596249740_GainShadows.png.f72f13da6611abfdedd787e1e5037e9f.png

 

But the important values are also the peak deltas, from medium to High could be 20!, and  from high to ultra 13!

 

Shadows.thumb.png.ecea309d7bd6b8911018269cdde94687.png

 

  • Like 1
  • Upvote 2

Share this post


Link to post
Share on other sites

 

 

Motheboard: Gigabyte Technology Co x570 Aorus Pro Wifi, Bios: F10a - AMD AGESA AM4 1.0.0.4

CPU: Ryzen 3800X

Core Speed: 4490.90Hz

L3 Cache: 2x16MB

RAM Type: DDR4

RAM Size: 64GB

RAM Freq 3592.8

RAM Latency: 16.0

GPU: 1080TI, Core clock +50, Memclock +500, Power 120%

STMark: 3044, 3056, 3013, 3025, 3106

*I'm actually surprised at these Single Thread results, because when I was running them about a month ago, I was typically getting in the mid-upper 2900's

 

High with Map Scenery distance unlimited:

Frames, Time (ms), Min, Max, Avg
 23613,    180000,  84, 231, 131.183

 

High with Map Scenery Distance 50km:

Frames, Time (ms), Min, Max, Avg
 23548,    180000,  87, 232, 130.822
 

High with Map Scenery distance Unlimited, and Distant Landscape detail at x4

Frames, Time (ms), Min, Max, Avg
 24069,    180000,  92, 234, 133.717

 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, and Horizon Draw distance at 150km

Frames, Time (ms), Min, Max, Avg
 23045,    180000,  86, 228, 128.028
 

 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on

Frames, Time (ms), Min, Max, Avg
 23563,    180000,  91, 225, 130.906
 

 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on

Frames, Time (ms), Min, Max, Avg
 23360,    180000,  89, 225, 129.778
 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme

Frames, Time (ms), Min, Max, Avg
 22820,    180000,  89, 226, 126.778
 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium

Frames, Time (ms), Min, Max, Avg
 25005,    180000,  93, 232, 138.917
 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple

Frames, Time (ms), Min, Max, Avg
 25419,    180000,  95, 237, 141.217
 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple, Grass Normal

Frames, Time (ms), Min, Max, Avg
 25254,    180000,  94, 233, 140.300
 

High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple, Grass Ultra

Frames, Time (ms), Min, Max, Avg
 24872,    180000,  94, 235, 138.178
 

HP Reverb with ASF off, High with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple, Grass Ultra

Frames, Time (ms), Min, Max, Avg
  8773,    180000,  40,  91, 48.739

 

Given that that ends up being pretty much the same results I got from the Quick mission I'd run earlier, I'm thinking this is good for getting VR performance as well.

The mission setup I had in that one was full cloud cover, 4x B-25 AI, 4x HE-111 AI, 4x P-47 (Player), 4x FW-190A-8 AI on the Rhineland map.

 

Addendum: Additional Tests

 

Ultra with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple, Grass Ultra

Frames, Time (ms), Min, Max, Avg
 23818,    180000,  92, 229, 132.322

 

 

HP Reverb with ASF off, Ultra with Map Scenery distance Unlimited, Distant Landscape detail at x4, Horizon Draw distance at 150km, SSAO on, HDR on, Clouds Extreme, Shadows: Medium, Reflections simple, Grass Ultra

Frames, Time (ms), Min, Max, Avg
  8252,    180000,  37,  69, 45.844

 

Setting it to Ultra does cause a hit, but, for me at least, the view is so worth it.

Edited by Voyager
Added test at Ultra detail

Share this post


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

STMark: 3044, 3056, 3013, 3025, 3106

 

That´s strange. Normally the differences are only 0 to 5 when re-running the STMark. But in your case the are too big from 3013 to 3106.

The reason could be that your CPU freq is not kept constant. Perhaps CPU throttling? or variable OC handled by the turbo tech of AMD?

3 hours ago, Voyager said:

High with Map Scenery distance unlimited:

Frames, Time (ms), Min, Max, Avg
 23613,    180000,  84, 231, 131.183

 

High with Map Scenery Distance 50km:

Frames, Time (ms), Min, Max, Avg
 23548,    180000,  87, 232, 130.822

 

I see that Map Scenery Distance has almost no influence in performance.

I didn´t specify that in the instructions, but in any case I was running that at 20Km which I think is the default value (I have never touched that).

Jason was explaining what is the purpose of that parameter:

https://forum.il2sturmovik.com/topic/34628-game-version-3001-discussion-final-release-of-the-battle-of-kuban/?do=findComment&comment=585554

 

Share this post


Link to post
Share on other sites

My run.  Note I turned off Lefuneste's mod, didn't see that in the instructions.

 

Motheboard: ROG Strix Z390-H

CPU: 9600k

Core Speed: 5 GHz

L3 Cache: 9 MB

RAM Type: DDR4

RAM Size: 16GB

RAM Freq 3200 MHz

RAM Latency: 16.0

GPU: 1080

STMark: 3016

 

Frames: 22942   - Avg: 127.456 - Min: 93 - Max: 222

Share this post


Link to post
Share on other sites

@chiliwili69 I believe it is the Windows Scheduler having a it's fun with things. Each core has a different performance peak, and the scheduler likes to shuffle the cores things run on.

 

Thank you for clearing up what the scenery distance handles. That does explain the 20gb charge I've been seeing. I'll have to try it with the ultra preset to see about getting more view distance. That does leave me wondering what the special setting for longer building render distance via multi threading was, though. 

 

Addendum: Looks like it is the 'Distant Buildings" setting in the settings options:

 

I've added two Ultra test results to my prior tests. It does seem to have a noticeable impact in VR, but I like the additional view distance enough that I'm going to keep it.

Edited by Voyager

Share this post


Link to post
Share on other sites
20 minutes ago, 71st_AH_statuskuo said:

My run.  Note I turned off Lefuneste's mod, didn't see that in the instructions.

 

Oops! Chili you should be aware I had Lefuneste's mod installed for my benchmark results. Sorry about that, it will throw things off a bit I guess.

Share this post


Link to post
Share on other sites
7 hours ago, 71st_AH_statuskuo said:

My run.  Note I turned off Lefuneste's mod, didn't see that in the instructions.

 

Yes, thanks. I should indicate in the instructions that the test should be run with Mods off.

I have been looking to the avg numbers achieved by people running lower cards than mine 1080Ti, for example 1060, 2070 and 1080.

They are achieved less fps than expected value and this could mean that the GPU is limiting the results in some periods of the test.

 

I also observed that clouds doesn´t load the CPU but the GPU. And the instructions said to run the benchmark with High clouds, which I think will limit the ability of lower cards (1060, 2070, 1080, 1070) to run "properly" the test. The purpose of the test was to bottleneck the CPU not the GPU.

 

So, I am thinking to change the instructions and relax the setting of the clouds to Low, since clouds almost has no impact in CPU.

 

People with higher cards will almost see not change in the test, but people with lower cards will see a fps gain.

 

Before doing that, could anybody of the previous people with lower cards than mine run again the test but setting the clouds to Low. (whitav8, AHstatuskuo, Stolle, Redwo1f). Thanks.

Share this post


Link to post
Share on other sites

As per request - benchmark settings as first post requirements, but clouds now set on LOW:

 

Frames: 15741 Time: 180000ms - Avg: 87.450 - Min: 53 - Max: 178

 

....................................................................

 

Just an aside as I have been playing with various settings combo's on my own as well, noticing something strange regarding clouds visually on my system. I normally play on HIGH general settings and have clouds set to HIGH as well, but if I keep clouds on HIGH and drop down to BALANCED settings (everything else being equal) - I notice a distant cloud shimmering at or near the horizon - yet this disappears simply by changing to the HIGH general settings. So there is something in the general config for HIGH not present in the BALANCED that fixes the problem in this regard. Indeed I have noticed that the only way for me to get rid of this horizon cloud shimmering is to have the clouds set to HIGH or EXTREME and be playing within the HIGH or ULTRA overall preset.

 

Anybody know (or have a good guess) what change to the HIGH preset config not present in the BALANCED suddenly fixes the HIGH detail horizon cloud shimmering for me?

 

 

Edited by Redwo1f

Share this post


Link to post
Share on other sites

Frames with High Clouds: Avg: 133.750 - Min: 95 - Max: 232
Frames with Low Clouds: Avg: 141.878 - Min: 98 - Max: 238

 

Motherboard: MSI MAG Z390 Tomahawk
CPU:                 9600K
CPU Freq:        5.0 Ghz
L3 cache:        9 MB
RAM type:        DDR4
RAM size:        16 GB
RAM Freq:        3200 MHz
RAM Latency: 16
GPU:                 2070
STMark:           3011

 

Share this post


Link to post
Share on other sites
16 hours ago, Redwo1f said:

As per request - benchmark settings as first post requirements, but clouds now set on LOW:

 

Frames: 15741 Time: 180000ms - Avg: 87.450 - Min: 53 - Max: 178

 

....................................................................

 

Just an aside as I have been playing with various settings combo's on my own as well, noticing something strange regarding clouds visually on my system. I normally play on HIGH general settings and have clouds set to HIGH as well, but if I keep clouds on HIGH and drop down to BALANCED settings (everything else being equal) - I notice a distant cloud shimmering at or near the horizon - yet this disappears simply by changing to the HIGH general settings. So there is something in the general config for HIGH not present in the BALANCED that fixes the problem in this regard. Indeed I have noticed that the only way for me to get rid of this horizon cloud shimmering is to have the clouds set to HIGH or EXTREME and be playing within the HIGH or ULTRA overall preset.

 

Anybody know (or have a good guess) what change to the HIGH preset config not present in the BALANCED suddenly fixes the HIGH detail horizon cloud shimmering for me?

 

 

 

You have phrased your observation on clouds with balanced very elegantly. My uncivilized method is saying clouds with balanced look like crap. 

 

I believe the reason is simply the result of what the different presets do. They increase/decrease the distance of higher quality rendered environment around you. With balanced, this 'good looking bubble' is so small that the mid range clouds fall out. If you go closer they stop looking like crap.

 

Whereas with high, the bubble is big enough to cover the mid range clouds, and high range clouds shimmering and pixelating is much less noticeable so overall it looks nice.

  • Upvote 1

Share this post


Link to post
Share on other sites

The Devs have never really explained what the presets really do. The simple explanation is that it increases the tree rendering line "bubble", but I have played enough with Balanced/High/Ultra to notice it is more than that. The cloud rendering on the horizon is one example, but there is also an improvement with textures which is noticeable when you look at the wing of your plane from inside the cockpit.

 

 

Share this post


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

The Devs have never really explained what the presets really do. The simple explanation is that it increases the tree rendering line "bubble", but I have played enough with Balanced/High/Ultra to notice it is more than that. The cloud rendering on the horizon is one example, but there is also an improvement with textures which is noticeable when you look at the wing of your plane from inside the cockpit.

 

When I last fiddled with it, I also felt that the preset changed the amount of detail in the cockpit geometry, but I wouldn't swear to it. It would make sense, though -- that's a lot of geometry right up close to you, dropping a few polygons from that would make it perform better.

Share this post


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

 

When I last fiddled with it, I also felt that the preset changed the amount of detail in the cockpit geometry, but I wouldn't swear to it. It would make sense, though -- that's a lot of geometry right up close to you, dropping a few polygons from that would make it perform better.

 

change it to Low and the cockpit will look horrible. I think there's plenty of additional cockpit fidelity with Ultra just based on that Low test.. 

 

but maybe I'm wrong.

 

Share this post


Link to post
Share on other sites
On 12/1/2019 at 9:57 PM, Voyager said:

High with Map Scenery distance unlimited:

Frames, Time (ms), Min, Max, Avg
 23613,    180000,  84, 231, 131.183

 

High with Map Scenery Distance 50km:

Frames, Time (ms), Min, Max, Avg
 23548,    180000,  87, 232, 130.822
 

High with Map Scenery distance Unlimited, and Distant Landscape detail at x4

Frames, Time (ms), Min, Max, Avg
 24069,    180000,  92, 234, 133.717

 

Thanks Voyager for your multiple test, but the first three test show something a bit unconsistant.

You get more fps with DistantLandscape at x4 than at x3 which is the setting for the test. It should not like that. You should achieve at least the same or lower, but not higher fps.

I think the reason is your variable turbo technology which is changing the CPU clock speed during the test.

You should try to find an overclock method to fix the CPU freq to a constant value (like we do it with intels), and then just repeat 3 or 4 runs with the exactly the same settings (the tests settings for example) and be sure that you get similar avg fps, normally a margin of noise of 1 fps up or down it could be ok, but not more.

On 12/2/2019 at 4:58 PM, Redwo1f said:

Frames: 15741 Time: 180000ms - Avg: 87.450 - Min: 53 - Max: 178

 

Many thanks. That´s confirm my theory. The test with lower  GPUs (like your 1060) was limited by GPU. You get 12 fps more now.

So I will change the instructions to run the test with Clouds at Low.

Anyone who already run the test before is welcome to run it again with  clouds low, just to have a results table with same settings.

On 12/2/2019 at 9:32 PM, pusicato said:

Frames with High Clouds: Avg: 133.750 - Min: 95 - Max: 232
Frames with Low Clouds: Avg: 141.878 - Min: 98 - Max: 238

 

Thanks. This again confirm the theory that clouds load the GPU and bottleneck the test for lower cards (2070 in your case) your delta was 8fps, smaller than the Redwo1f comparison since your card is a 2070.

Share this post


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

The cloud rendering on the horizon is one example, but there is also an improvement with textures which is noticeable when you look at the wing of your plane from inside the cockpit

 

The cloud rendered is because the detailed render bubble is bigger or smaller.

But I remember there were also some old post were it was explained that the Presets changed the number of polygons of every objects. You can see that in the screew of the plane if you look to them closely. So it results in a more complex scene to render , larger and with more polygons. That´s why it loads both, CPU and GPU.

Share this post


Link to post
Share on other sites
On 12/1/2019 at 7:39 PM, chiliwili69 said:

 

WOW! what a nice collection of rig. Al  of them would be interesting, specially the Rig1 and Rig2. You can try to disable some cores of the CPU to low temp and reach higher freq. IL-2 doesn´t need more than 4 cores.

 

 

Well, i don't think I am limited by temps, I have some thermal headroom on all oht the rigs. Especially with the 9700K, it seems I got super lucky in the sillicon lottery. MB prediction says it could do 5.5 ghz AVX with my cooler. I don't really belive that, but:

 

https://valid.x86.fr/izw47p

- see the voltage and temps. It is completely stable and i have never seen temps exceeding 70 degrees (voltage is on auto). Its really strange.

 

BTW, if anybody wants to ask anything about that hardware, comparision or something, just ask...

Share this post


Link to post
Share on other sites

I ran the benchmarks exactly as Chili instructed (clouds high) and these were the results. Specs in my sig

 

CPU Mark: 2496

2019-12-04 22:10:04 - Il-2

Frames: 21582 - Time: 180000ms - Avg: 119.900 - Min: 80 - Max: 259

 

These results seem pretty competitive based on the scores I see from others. The max fps was one of the fastest I've seen here which really surprised me. I have been considering an upgrade to a 9700k based rig because my current setup on Ultra with a few settings backed down a bit seems to struggle some when things get busy. Im looking at trying vr (probably Rift S) and I thought I'd need to upgrade to really enjoy it.

 

How should I interpret these results in terms of my current components? Do they suggest any issue with CPU, RAM etc?

 

 

Edited by TheSNAFU

Share this post


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

These results seem pretty competitive based on the scores I see from others. The max fps was one of the fastest I've seen here which really surprised me. I have been considering an upgrade to a 9700k based rig because my current setup on Ultra with a few settings backed down a bit seems to struggle some when things get busy. Im looking at trying vr (probably Rift S) and I thought I'd need to upgrade to really enjoy it.

 

How should I interpret these results in terms of my current components? Do they suggest any issue with CPU, RAM etc?

 

Many of my squadmates are running a Rift S + GTX 1080, and they have simply decided to live with 40 FPS + ASW mode to interpolate up to 80 FPS. The Rift tech really works surprisingly well here. I think your rig would do fine for this, but it would struggle (both CPU and GPU) to get to a real 80 FPS. While my squadmate (and I) are fairly happy with ASW, some pilots really don't like it at all and do everything they can to get to a real 80 FPS (or 90, depending on headset). Are there any places for you to buy a Rift S with money back guarantee? You could always return it if you don't like it.

Share this post


Link to post
Share on other sites

Did a quick and dirty overclock at 5.2ghz on my 9600k 

 

Passmark ST 3180

 

Hopefully will have time to do full benchmark at the weekend

 

Cheers, Dakpilot 

Share this post


Link to post
Share on other sites
On 12/5/2019 at 12:59 AM, thermoregulator said:

- see the voltage and temps. It is completely stable and i have never seen temps exceeding 70 degrees (voltage is on auto). Its really strange

 

Hey!! from what planet is your CPU!! that´s pretty amazing, you reach 5.2 with just 0.98voltage!!. You can try to go directly to 5.4 with 1.2v watchingout temps.

I am very interested in what OC you can really reach and what results do you have in the becnhmark at 5.2 and higher freqs.

 

Also, you said that your MB prediction is 5.5AVX (on or off?). From where do you see that prediction?

On 12/5/2019 at 4:47 AM, TheSNAFU said:

I ran the benchmarks exactly as Chili instructed (clouds high)

 Thank you for running the test, although I still have to change the instructions and the image of the settings since the base test has been changed to clouds low, but additionally I also asked to run the test with clouds High.

 

Regarding your CPU, you STMark,  your RAM at 1600 (I don´t know latency) you achieve and incredible max fps. And a good avg fps. The 259 fps for the max is really strange. Normally I could find reason for people who obtain low performance, but it is difficult to find the reasons for people who obtain a higher performance (like Alonzo with their CPU freq test).

Please, if you have time could you please check that you set the same graphics settings (please be sure that you run at 1920x1080 since I saw that your monitor is an HD TV, not a FHD TV)

and repeat 2 or 3 times the test with clouds high and low. We need to find plausible explanations to every fact if we want to extract some knowledge.

 

With your verification we could then see if your system could be valid for VR with the Rift-S. But in anycase, you can buy the Rift-S and upgrade later if you see that performance is not good.

Share this post


Link to post
Share on other sites
On 12/5/2019 at 6:36 AM, Dakpilot said:

Did a quick and dirty overclock at 5.2ghz on my 9600k 

 

Passmark ST 3180

That´s a very nice STMark. Let´s see how it goes. If you have time, try to run the test at other frequencies like 4.0 and 4.5 Ghz. Remember the clouds to low. Thanks.

Share this post


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

 

Hey!! from what planet is your CPU!! that´s pretty amazing, you reach 5.2 with just 0.98voltage!!. You can try to go directly to 5.4 with 1.2v watchingout temps.

I am very interested in what OC you can really reach and what results do you have in the becnhmark at 5.2 and higher freqs.

 

 

 

Sounds more like it is using an adaptive voltage.

The key would be seeing what is being applied under load.

If I set my voltage to adaptive, I can set my bios vcore around .95. Yet under load my 9900k will be pumping out 1.34-1.38v VID.

In bios with it set to that the voltage will read around 1.35v.

So I would be very careful applying much higher voltage in the bios, make sure and monitor what is really being applied. Just throwing in 1.2v without knowing how much more is being applied under load could be risky.

 

On my EVGA Z390 Dark board, I can set Vcore either Auto, Adaptive, or Override. I do not care for auto as it applies more than is really needed. Adaptive has the behavior described above. I personally prefer the Override, as I can set it more to what it actually will run.

Mine is set to 1.29v (Override) with my 5.1 Ghz all core overclock, and it will read around 1.34-1.36v whist stress testing. This option gives me the better control of my vcore.

Obviously different boards will have different settings/nomenclatures.

I would not take a board that is stable with a vcore at .98v and pump it up immediately to 1.2v without understanding first what voltage it is actually applying under load.

Edited by dburne

Share this post


Link to post
Share on other sites

Chili, thanks for taking a look at my rigs scores and helping me understand the results. I ran the benchmark 3 times with high clouds and 3 with low clouds. Scores are below. It doesn't look like there is much difference between low and high, in fact max fps looks faster at high. Seems weird to me that this pc is doing these scores. I verified all game and Nvidia settings are as you instruct. Game and monitor resolution is 1920x1080 native for the HDTV.

 

High Clouds


2019-12-06 20:12:24 - Il-2
Frames: 23187 - Time: 180000ms - Avg: 128.817 - Min: 75 - Max: 261

2019-12-06 20:19:08 - Il-2
Frames: 21492 - Time: 180000ms - Avg: 119.400 - Min: 79 - Max: 256

2019-12-06 20:29:22 - Il-2
Frames: 21983 - Time: 180000ms - Avg: 122.128 - Min: 79 - Max: 260


Low Clouds

2019-12-06 20:38:22 - Il-2
Frames: 22660 - Time: 180000ms - Avg: 125.889 - Min: 81 - Max: 257

2019-12-06 20:47:27 - Il-2
Frames: 22156 - Time: 180000ms - Avg: 123.089 - Min: 75 - Max: 255

2019-12-06 20:51:45 - Il-2
Frames: 22390 - Time: 180000ms - Avg: 124.389 - Min: 79 - Max: 252

 

Edited by TheSNAFU

Share this post


Link to post
Share on other sites

Thanks for the standardized benchmark, @chiliwili69!

 

Although you ruined my day since my rig in this benchmark aims at the bottom of the table 😂. And rightly so, to be honest, since it is an old rig with a top-class GPU (it is an intermediary step to a new rig sometime next year 😎; under my "normal" settings it averages 130fps, ufuf!).

 

Motherboard: ASUS Z97-PRO(Wi-Fi ac)
CPU:                 i7 4790K
CPU Freq:        4.5 Ghz
L3 cache:        8 MB
RAM type:       DDR3
RAM size:       16 GB
RAM Freq:      1600 MHz
RAM Latency: 9
GPU:                2080Ti
STMark:          2687
 
FRAPS TXT:
2019-12-08 16:45:51 - Il-2
Frames: 19869 - Time: 180000ms - Avg: 110.383 - Min: 69 - Max: 200

 

BOX  4.002

 

I also attach EN version of the nvidia control panel.

 

Milan

 

Global Settings 1.jpg

Global Settings 2.jpg

Program Settings.jpg

Edited by apollon01

Share this post


Link to post
Share on other sites
On 12/6/2019 at 12:02 PM, dburne said:

So I would be very careful applying much higher voltage in the bios, make sure and monitor what is really being applied. Just throwing in 1.2v without knowing how much more is being applied under load could be risky.

 

Good advice Don. I didn´t know about the adaptive voltage. I thought it was a fixed value, that´s why it was so strange.

For overclock I also do as you indicate, using fixed custom voltage (override mode).

On 12/7/2019 at 3:04 AM, TheSNAFU said:

I ran the benchmark 3 times with high clouds and 3 with low clouds. Scores are below. It doesn't look like there is much difference between low and high, in fact max fps looks faster at high. Seems weird to me that this pc is doing these scores. I verified all game and Nvidia settings are as you instruct. Game and monitor resolution is 1920x1080 native for the HDTV.

 

Many thanks for running again the tests with the new settings (Clouds Low).

It is clear that your 1080 GPU was not bottlenecking the test too much, since the average results of the 3 tests does not differ too much.

Yes, your PC delivers a quite nice result. This is a fact. And I really don´t know why, but I took a closer look at the specs of the i7-3820 CPU and I saw it is a CPU of 4 channels RAM, allowing a higher max memory bandwidth:

 

i7-3820.png.bc80684c80a3ac2277c95a7b10b16466.png

 

It is the first time we test a CPU like that. So, perhaps this could be the reason.

If this is like that we could search for CPUs with that capability.

 

Share this post


Link to post
Share on other sites

These are the intel CPU with 4-channel memory, :

 

i9-10980XE
i9-10940X
i9-10920X
i9-10900X
i9-9900X
i7-9800X
i9-7900X
i7-7820X
i7-7800X
i7-6950X
i7-6900K
i7-6850K
i7-6800K
i7-5960X
i7-5930K
i7-5820K
i7-4960X
i7-4930K
i7-4820K
i7-3970X
i7-3960X
i7-3930K
i7-3820


Is anyone here with that CPUs to test the benchmark?

 

Share this post


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

Although you ruined my day since my rig in this benchmark aims at the bottom of the table 😂. And rightly so, to be honest, since it is an old rig with a top-class GPU (it is an intermediary step to a new rig sometime next year 😎;

 

LOL!  😂

Thank you for your test. All data is valuable. You have the same CPU than me with the topline GPU available (2080Ti), but this test shows how important is the CPU and RAM in this game.

As expected, your result is along the lines of STMark index. I also plan next year to upgrade CPU/RAM so I am interested to gain as much as info from what is important for this game.

Thank you for the pics in english.

Edited by chiliwili69

Share this post


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

 

Good advice Don. I didn´t know about the adaptive voltage. I thought it was a fixed value, that´s why it was so strange.

For overclock I also do as you indicate, using fixed custom voltage (override mode).

 

 

Even with fixed voltage (override) under load the motherboard will typically throw some additional voltage to the CPU that it thinks it needs, depending on the LLC (vdroop) set in bios. But it is much better controlled and reasonable unlike it is with an adaptive voltage setting.

I like to monitor my Vcore (rather than vid) whilst stress testing the system. For my 9900k I prefer it not to exceed 1.4v under heavy load.

Gaming it is certainly less.

For my 9900k at 5.2 GHz my vcore is typically around 1.36v under heavy load. I have it set to 1.318v in the bios. Under load my core temps are around 70-73c whilst stress testing. Upper 60's typically while gaming.

 

I tried increasing to 5.3 GHz, however the additional voltage I was having to throw at it to try and get it stable during a stress test I was uncomfortable with, so have settled at least for now at 5.2.

Edited by dburne

Share this post


Link to post
Share on other sites

HEY!! taking now a closer look of previous test (Balapan old test in VR) I have just picked a CPU processor (the 6800K) with 4 channel RAM. It was 18 fps above the expected correlation.

6800K.thumb.png.a80f9d050c7a4d883aea4e07dcbf6628.png

 

It would be good to have more tests of the latest 4-channel CPUs to confirm how they help for monitor and VR.

Share this post


Link to post
Share on other sites

Ok I ran the benchmark. All settings exactly as requested in the OP by Chili.

 

Motherboard: EVGA Z390 Dark
 CPU:                 i9 9900k
 CPU Freq:        5.2 GHz
 L3 cache:        16 MB
 RAM type:        DDR4
 RAM size:        32 GB
 RAM Freq:        3200 MHz
 RAM Latency: 14
 GPU:                 EVGA 2080 Ti FTW3
 STMark:           3154

 

Fraps Log:

2019-12-09 10:14:06 - Il-2
Frames: 28948 - Time: 180000ms - Avg: 160.822 - Min: 111 - Max: 268

 

 

 

Passmark.jpg.ae841fd6a3e700059ed13fb22213ad13.jpg

 

Edited by dburne

Share this post


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

Avg: 160.822 - Min: 111 - Max: 268

 

Many thanks Don for running the test. You have a top performer machine!! You are 7fps above expected.

 

Just curiosity, Have you tried to run your RAM at other higher Intel XMP profiles? Just to see if there is an impact.

Share this post


Link to post
Share on other sites
On 11/29/2019 at 4:28 PM, wju said:

**************************************************************************************

Motherboard: MSI X299 Gamming Pro Carbon AC
 CPU:             i7 7820x
 CPU Freq:        4.7 Ghz  AVX=0,AVX512=0
 L3 cache:        11 MB
 RAM type:        DDR4
 RAM size:        32 GB
 RAM Freq:        3466 MHz
 RAM Latency:         16
 GPU:                 1080Ti MSI, max. core 2015Mhz
 STMark:           2703
 
FRAPS TXT:
2019-11-29 16:02:11 - Il-2
Frames: 22324 - Time: 180000ms - Avg: 124.022 - Min: 82 - Max: 215

 

BOX  4.002

 

wju, I have just realized that your CPU is a 4-channel RAM chip!

I really want to know more about your machine:

1.- are you actually using 4-channel? (ie is your RAM config 2x16 or 4x8?)

2.- What CPU cooler are you using?

3.- have you tried to Overclock more by putting Hyperthreading Off and/or disabling 2 or 4 cores of your 8 physical cores? (IL-2 run well with just 4 cores)

4.- could you run again the benchmark with the new settings? ( Clouds  to Low to not bottleneck GPU)

 

Thanks!

 

Edited by chiliwili69

Share this post


Link to post
Share on other sites


Well, I have just re-run the benchmark with the new settings (Clouds Low) and also tested 3 different frequencies making 3 runs per each freq:


freq 4.0GHz

Frames: 19650 - Time: 180000ms - Avg: 109.167 - Min: 73 - Max: 190

Frames: 19960 - Time: 180000ms - Avg: 110.889 - Min: 74 - Max: 189

Frames: 20173 - Time: 180000ms - Avg: 112.072 - Min: 74 - Max: 196

freq 4.4GHz

Frames: 21755 - Time: 180000ms - Avg: 120.861 - Min: 81 - Max: 212

Frames: 21227 - Time: 180000ms - Avg: 117.928 - Min: 75 - Max: 206

Frames: 21515 - Time: 180000ms - Avg: 119.528 - Min: 80 - Max: 204

freq 4.8GHz

Frames: 22909 - Time: 180000ms - Avg: 127.272 - Min: 86 - Max: 215

Frames: 22894 - Time: 180000ms - Avg: 127.189 - Min: 85 - Max: 215

Frames: 22879 - Time: 180000ms - Avg: 127.106 - Min: 85 - Max: 213

 

I have compared these new results with my previous runs with Clouds High:

 

Freq 4.0GHz

Clouds Low: 110.7 (avg of 3 runs)

Clouds High: 111.1

 

Freq 4.4GHz

Clouds Low: 119.4 (avg of 3 runs)

Clouds High: 119.6

 

Freq 4.8GHz

Clouds Low: 127.2 (avg of 3 runs)

Clouds High: 125.2

 

So, for lower freqs (4.0 and 4.4) my 1080Ti was not bottlenecking the test since I obtain almost similar results.
But at 4.8 my 1080Ti was constraining the test a little bit.

Now (with clouds low) the delta between 4.0 and 4.4 is about 8.5 fps, and the delta between 4.4 and 4.8 is about 8 fps.

 

I have moved all previous tests with High clouds to the bottom of the table, to avoid mixing results performed at different settings.

Share this post


Link to post
Share on other sites

Motherboard: Gigabyte X570 Aorus Master
CPU: AMD     Ryzen 5 3600x
cpu freq:    4400.55MHz
L3 cache:    2x16 MB
Ram type:    DDR4
Ram size     32 GB
Ram speed:   3600MHz
Ram latency:  16
GPU:        1080Ti
STMark:     3115

 

Low settings:

2019-12-09 15:07:47 - Il-2
Frames: 28548 - Time: 180000ms - Avg: 158.600 - Min: 104 - Max: 263

 

High settings: 

 2019-12-09 14:55:19 - Il-2
Frames: 24984 - Time: 180000ms - Avg: 138.800 - Min: 92 - Max: 237

 

Ultra settings:

2019-12-09 15:26:35 - Il-2
Frames: 23721 - Time: 180000ms - Avg: 131.783 - Min: 89 - Max: 236

 

 

Edited by Jaws2002

Share this post


Link to post
Share on other sites
20 minutes ago, Jaws2002 said:

On High:   2019-12-09 14:55:19 - Il-2
Frames: 24984 - Time: 180000ms - Avg: 138.800 - Min: 92 - Max: 237

 

Thanks for running again. As expected, with your 1080Ti there is almost no difference in fps with respect your previous test.

 

But your STMark in the previous run was 3040 and now 3115 both at the same CPU freq

Edited by chiliwili69

Share this post


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

 

Many thanks Don for running the test. You have a top performer machine!! You are 7fps above expected.

 

Just curiosity, Have you tried to run your RAM at other higher Intel XMP profiles? Just to see if there is an impact.

 

No I leave my ram at it's XMP of 3200 MHz CL14.

Share this post


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

 

Thanks for running again. As expected, with your 1080Ti there is almost no difference in fps with respect your previous test.

 

But your STMark in the previous run was 3040 and now 3115 both at the same CPU freq

 

Last time i did the test just once. Now i got similar score the first run, but i also did a rerun and it gave me better score, so i logged that one.:lol:

 

 

efsgfsfs.png

Edited by Jaws2002

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...