Jump to content

2016 Hardware Benchmark - DCS World 1.5.x


Recommended Posts

Yes, there is actually! Download and install the Oculus runtime.

https://developer.oculus.com/

 

There's a setting somewhere that lets it run without the headset for development purposes. You will still get an image in a window on your monitor. It will not look great, but it might work for the purpose of your tests. I don't remember exactly what the setting was called, but you will probably find it on your own. Maybe it was "debug mode" or something?

 

Thanks, I'll try to figure it out :)

 

Are you still planning a benchmark for the R9 295x2?

Yes, for sure, probably later today or over the weekend. You guys are lucky, the weather is bad, lots of fresh snow, climbing is off for at least this weekend. I don't want to try out the avalanche equipment (again)... Once was enough :doh:

Link to comment
Share on other sites

  • Replies 235
  • Created
  • Last Reply

Top Posters In This Topic

AMD cards are great but the thing is DCS performs better on NVIDIA cards at the moment. Truth be told, DCS engine is fairly aged (not that it hasn't aged well) and AMD never bothered to fix somewhat broken legay functions (including DX9) performance when they purchased ATi back in 2006.

 

Wait? You are making these tests on version 1.5, right? 1.5 introduced an entirely new graphics rendering engine based on dx11. As an AMD user (290X) I have to say the game went from unplayable (1.2.16 dx9) to running really well (1.5 dx11) while looking much better at the same time.

Link to comment
Share on other sites

Yes, the test are being conducted on 1.5.2, correct. DCS 1.5 engine is indeed light years better than 1.2!

 

I am trying to say the legacy functions, much of them carried over from older standard, doesn't seem to run as good as on NVIDIA. The proof to this are the new games, built on entirely new engines, where AMD cards more than adequately keep up the pace with NVIDIA.

 

I hope I'm not causing too much confusion.


Edited by tiborrr
Link to comment
Share on other sites

what kind of cooler you got running?

 

corsair H80

 

P.S. there are nvidia fanboys too... the secret is to use what's best, for what games you play, not to use what colour, name you like more, or what works best in games you don't play. nothing wrong or insulting about being a fanboy, unless you think it's bad to be a fan of something... except if you're going to say what you have is better non-objectively, that makes you the wrong kind of fanboy.

 

the objective here is getting DCS running well, right? so objectively nvidia is better than ati, according to these graphs.

 

anyway, the fact that a 3 year old card is still beating brand new high end cards is lolish... shows how much progress has slowed on gpu and cpu speeds... amd, intel and nvidia.. the 780 and sandy were awesome investments when they were new.


Edited by Hadwell

My youtube channel Remember: the fun is in the fight, not the kill, so say NO! to the AIM-120.

System specs:ROG Maximus XI Hero, Intel I9 9900K, 32GB 3200MHz ram, EVGA 1080ti FTW3, Samsung 970 EVO 1TB NVME, 27" Samsung SA350 1080p, 27" BenQ GW2765HT 1440p, ASUS ROG PG278Q 1440p G-SYNC

Controls: Saitekt rudder pedals,Virpil MongoosT50 throttle, warBRD base, CM2 stick, TrackIR 5+pro clip, WMR VR headset.

[sIGPIC][/sIGPIC]

Link to comment
Share on other sites

Tiborr, that's some absolutely awesome work you have done.

 

One request though, could you please attach the used .trackfiles to the first post?

This way people can run the same benchmark, compare results and maybe even identify bottlenecks on there systems.

 

Looking forward to the Nevada tests.

[sIGPIC][/sIGPIC]

The keeper of all mathematical knowledge and the oracle of flight modeling.:)
Link to comment
Share on other sites

Hello everybody,

 

ran a benchmark with GTX 960 and R9 295X2 while the NTTR is downloading, thanks to Decibel_dB! :)

 

Initial finding: Dang, we have the new best bang for a buck card for WQHD resolution. If the money is tight and you want the latest gen GPU, the GTX 960 might the right choice. Crossfire however has micro-stutter issues, I have read about it on these forums and there only fix for it seem to be to set the MODEL VISIBILITY to OFF. However, I haven't found that function anywhere in the menu.

 

Regarding 'VR' testing:

I have found a way to sort-of simulate VR using Stereo profile inside DCS while setting a custom resolution of 2160x1200 (2x 1080x1200 like the Rift). The initial results are pretty similar to 3 Monitor "High FOV" profile, but the performance drop is not so steep. I have to retest it using several different GPUs of various performance range to see whether it is system/DCS engine bound or GPU bound. I have a feeling it's again the engine...

 

Back to regular FHD, WQHD and 4K results:

 

Eq4jNv2.png

 

xZTxHxt.png

 

qOJlWkW.png

 

 

The TL;DR sections has also been updated (see #1 post):

200_s.gif

smartass.gif What have we learned so far (tl;dr):

Compiled list of lessions learned:

1. If budget allows - get a fast (overclockable) quad-core CPU! Yes, DCS scales great up to 4 cores!

2. Overclock your CPU: Frequency is the king! Most of the GPU performance will be CPU frequency bound.

3. Faster memory yields noticable performance gains, especially in the minimum FPS range

4. Flying in 4K on PC: On AMD R9 series - make sure to use DisplayPort (DP) 1.2 for 4K@60Hz - do not use HDMI as HDMI 1.4 isn't cutting it (only 30Hz refresh rate). On NVIDIA GTX 9xx series use either DP1.2 or HDMI (supports HDMI 2.0).

5. Flying in 4K on TV: AMD is currently not a good choice for 'cheap' TV 4K. The problem is that these R9 200/300/Fury/Nano cards can only put out 4K@30Hz over HDMI 1.4. The only way to get 4K@60Hz on current gen AMD is to use DisplayPort (DP) video outputs, which works great - 4K PC monitors with DP video inputs are readily available, however DP 4K TVs are not. If your 4K TV only supports HDMI it's better if you get NVIDIA. Alternatively you can purchase DP 1.2 to HDMI 2.0 adapter which bypasses all of the aforementioned issues. At the time of writing (Jan 2016) the cheapest 55" 4K@60Hz TVs with DP video inputs are still 8-10x more expensive than their HDMI 2.0 counterparts.

6. Overclocking of the GPUs tested yield noticable gains starting with WQHD resolution. Lower performance GPUs than the ones tested (e.g. GTX 750) will benefit from GPU overclocking even in low resolutions.

7. It is safe to say GTX 960 is currently the price/performance king for playing DCS at WQHD (2560x1440) while GTX 970 is the P/P king for playing FCS in 4K. For playing DCS at FHD resolution an AMD Radeon R7 370 (basically only about 10-15% slower than R9 270X) or NVIDIA GeForce GTX 950 will be great. If money is tight consider higher-end last gen GPUs.

8. CrossFireX work and scales nicely but seem to cause micro-stuttering. The latter can be remedied by setting MODEL VISIBILITY to off (however I haven't found that function in settings as of version 1.5.2).

9. Multi-monitor WALL setup - with 'regular' FOV (field-of-view, e.g. "1 Monitor" preset within the game settings) - performs equally good as the single-monitor setup at comparable (total) resolution. Therefore it's possible to build a cheap large, high-resolution screen out of smaller panels.

10. Multi-monitor BOX setup - with high FOV (field-of-view, e.g. "3 Monitor" preset within the game settings) - heavily taxes the FPS, regardless of your graphics card, most likely because of the DCS 1.5 engine limitation. A workaround is to set the MEDIUM graphics preset to enjoy high FPS.

The stuff that remains:

- VR resolution analysis

- CPU architecture testing

- DCS World 2.0 NTTR range benchmark

 

Best Regards,

Niko


Edited by tiborrr
  • Like 2
Link to comment
Share on other sites

Your result upset me so much , i have a 1920x1200 monitor for my computer and i`m feeling its such a waste , but i cant afford a QHD or 4K atm.

 

p.s

no GTX 980 tests?

IAF.Tomer

My Rig:

Core i7 6700K + Corsair Hydro H100i GTX

Gigabyte Z170X Gaming 7,G.Skill 32GB DDR4 3000Mhz

Gigabyte GTX 980 OC

Samsung 840EVO 250GB + 3xCrucial 275GB in RAID 0 (1500 MB/s)

Asus MG279Q | TM Warthog + Saitek Combat Pedals + TrackIR 5

[sIGPIC][/sIGPIC]

Link to comment
Share on other sites

What an awesome benchmark, thanks! As I'm loking for a new build, I'll monitor this topic with great interest.

 

Could you comment on the use of Hypertreading for DCS? I'm currently deciding on an i5 or i7 (K) processor.

Link to comment
Share on other sites

Excellent benchmarks and information for hardware updates!

It brings light to years of vague testing and speculation about CPU (cores and OC GHz), GPU and RAM effect on DCS FPS.

+1 Rep.

AKA TANGO-117. DCS Modules: ALL. I7 6700k @ 4.9 GHz / 32 GB DDR4 @ 3.2 GHz / 950 Pro m.2 + 4xSSDs / Gigabyte 1080TI 11 GB OC / 48" 4K Curved Samsung TV / TM Warthog Hotas / TM TPR rudder pedals / Track IR. Private pilot and sailplane pilot in RL.

Link to comment
Share on other sites

Hello everybody again!

 

@T_A:

Sadly, I'm not able to get GTX 980 at the moment. My guess it somewhere between the 980 Ti and the 970, closer to the latter. If you feel irritated just crank up the AA to make your card sweat a little bit :)

 

@Lenop:

Hyperthreading helps for sure (on an i3 for example), but there's no more performance gains going from 4cores to 8 "cores" (or threads). If you want to save money, get the i5 K, a good liquid cooler (like EK ;)) for the CPU and overclock that beast. Frequency and ICP are the kings once you go quad-core or better.

 

@JorgeIII:

Thanks man!

 

5.7 VR resolution gaming performance analysis:

The purpose of this test is to test out the impact of the running stereo "VR" 2160x1200 resolution display setup (thus simulating Oculus Rift VR headset) on the general FPS performance. The performance hit is not as obvious as on the "3 Monitor" High FOV display setup but still big enough to conclude the DCS engine is starving the cards. A 4K resolution results (HIGH preset) is added to the chart for reference.

 

HW setup:

- CPU: Intel Core i7 5775C @ 4GHz core / 3.3GHz uncore

- MB: ASUS ROG Maximus VII Gene

- RAM: 4x4GB DDR3-2133 C10 12-12-28 1T @ 1.35V

- GPU: GTX 960 and GTX 980 Ti

- Drive: 128GB Crucial BX100

- OS: Windows 10 Pro x64

- Cooling: EK-XLC Predator 240 - liquid cooled CPU; factory cooling on GPUs

- Monitor: Dell 2713HM

- Drivers: Nvidia 361.43

Image settings as tested (2160x1200 "Stereo" mode):

31xVw4Kl.jpg

 

Results:

ksVQnWc.png

 

Simulated VR resolution performance analysis:

- Ironically enough, GTX 960 (a 200€ card) here performs identically to the GTX 980 Ti (a 650€ card) which further points to an obvious bottleneck, which is most likely not hardware related. Since the GPUs are pushing a mere ~ 2.6MPix image here (compared to ~ 8.3MPix at 4K), both GPUs more than suffice for the workload.

- In order to maintain higher FPS the same tricks apply as when running "3 Monitor" High FOV setup. Drop the preset to either MEDIUM or LOW.

- Just like before, the GPUs are starving pretty much for the majority of the time when image quality is set to HIGH preset. The only time GPUs run at peak performance is when the view is facing towards the sky or clouds (no ground objects needs to be drawn). - I have also tested custom HIGH preset (only Visibility set to LOW and Trees set to MINIMUM) trying to show the effect of number of polygons/objects on the FPS.

- Doesn't appear to be CPU nor GPU bound. Has to be DCS engine limitation or simply GPUs cannot cope with number of polys on the screen.

 

I will now go onto CPU architecture testing. I have done some testing on the AMD setups but found out I'm loosing activation tickets like it's nothing. The system considers swapping out a CPU a major hardware change (while it didn't say anything when I swapped out a ton of GPUs) so I only have a few activations left - even though I deactivated them as per instructions :(

 

Best Regards,

Niko

 

P.S.: Also updated TL;DR, see #1 post:

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

 

200_s.gif

smartass.gif What have we learned so far (tl;dr):

Compiled list of lessions learned:

1. If budget allows - get a fast (overclockable) quad-core CPU! Yes, DCS scales great up to 4 cores!

2. Overclock your CPU: Frequency is the king! Most of the GPU performance will be CPU frequency bound.

3. Faster memory yields noticable performance gains, especially in the minimum FPS range

4. Flying in 4K on PC: On AMD R9 series - make sure to use DisplayPort (DP) 1.2 for 4K@60Hz - do not use HDMI as HDMI 1.4 isn't cutting it (only 30Hz refresh rate). On NVIDIA GTX 9xx series use either DP1.2 or HDMI (supports HDMI 2.0).

5. Flying in 4K on TV: AMD is currently not a good choice for 'cheap' TV 4K. The problem is that these R9 200/300/Fury/Nano cards can only put out 4K@30Hz over HDMI 1.4. The only way to get 4K@60Hz on current gen AMD is to use DisplayPort (DP) video outputs, which works great - 4K PC monitors with DP video inputs are readily available, however DP 4K TVs are not. If your 4K TV only supports HDMI it's better if you get NVIDIA. Alternatively you can purchase DP 1.2 to HDMI 2.0 adapter which bypasses all of the aforementioned issues. At the time of writing (Jan 2016) the cheapest 55" 4K@60Hz TVs with DP video inputs are still 8-10x more expensive than their HDMI 2.0 counterparts.

6. Overclocking of the GPUs tested yield noticable gains starting with WQHD resolution. Lower performance GPUs than the ones tested (e.g. GTX 750) will benefit from GPU overclocking even in low resolutions.

7. It is safe to say GTX 960 is currently the price/performance king for playing DCS at WQHD (2560x1440) while GTX 970 is the P/P king for playing FCS in 4K. For playing DCS at FHD resolution an AMD Radeon R7 370 (basically only about 10-15% slower than R9 270X) or NVIDIA GeForce GTX 950 will be great. If money is tight consider higher-end last gen GPUs.

8. CrossFireX work and scales nicely but seem to cause micro-stuttering. The latter can be remedied by setting MODEL VISIBILITY to off (however I haven't found that function in settings as of version 1.5.2).

9. Multi-monitor WALL setup - with 'regular' FOV (field-of-view, e.g. "1 Monitor" preset within the game settings) - performs equally good as the single-monitor setup at comparable (total) resolution. Therefore it's possible to build a cheap large, high-resolution screen out of smaller panels.

10. Multi-monitor BOX setup - with high FOV (field-of-view, e.g. "3 Monitor" preset within the game settings) - heavily taxes the FPS, regardless of your graphics card, most likely because of the DCS 1.5 engine limitation. A workaround is to set the MEDIUM graphics preset to enjoy high FPS.

11. VR setup (2160x120 - "Stereo" mode) is - despite the low resolution - very demanding on this game and one needs to resort to lowering quality settings in order to achieve 75FPS+. A high end and a low end GPU perform pretty exactly the same, a sign of engine bottleneck.


Edited by tiborrr
Link to comment
Share on other sites

 

I will now go onto CPU architecture testing. I have done some testing on the AMD setups but found out I'm loosing activation tickets like it's nothing. The system considers swapping out a CPU a major hardware change (while it didn't say anything when I swapped out a ton of GPUs) so I only have a few activations left - even though I deactivated them as per instructions :(

 

 

Contact customer support and they will gladly restock your activations.

You could point them to this thread to explain why your using up so many in such a short time if need be.

 

Thank's for the effort Tiborrr, much appreciated testing your doing.

[sIGPIC][/sIGPIC]

The keeper of all mathematical knowledge and the oracle of flight modeling.:)
Link to comment
Share on other sites

IMHO your setup is not capable of showing differencies beyond 100fps !?

 

Either your CPU or Monitor caps the fps around 100fps and thus no card can be pushed to its limits.

 

I have far higher fps at 1440p with a single 980, so I honestly doubt that chart, a 970 to Titan all have same fps ???? no way, only if you cap it somewhere else.

Gigabyte Aorus X570S Master - Ryzen 5900X - Gskill 64GB 3200/CL14@3600/CL14 - Asus 1080ti EK-waterblock - 4x Samsung 980Pro 1TB - 1x Samsung 870 Evo 1TB - 1x SanDisc 120GB SSD - Heatkiller IV - MoRa3-360LT@9x120mm Noctua F12 - Corsair AXi-1200 - TiR5-Pro - Warthog Hotas - Saitek Combat Pedals - Asus PG278Q 27" QHD Gsync 144Hz - Corsair K70 RGB Pro - Win11 Pro/Linux - Phanteks Evolv-X 

Link to comment
Share on other sites

BitMaster: Note that the ~100fps numbers are average framerate while playing back a specific track. Look at the max fps, and you see that it even approaches 200fps is some cases. Obviously there's no fps cap involved. I'm absolutely certain that if you played back the same track on your PC, you would see no more than around 100fps average either.

Link to comment
Share on other sites

Very interesting indeed. Thanks for doing this, it will help me a lot since I’m about to buy and build a new dedicated system for DCS.

 

So, this really confirms that 1 monitor option is preferred if you have multiple monitors. Nice to have some hard facts, and not just opinions.

 

Hopefully the final 2.0 engine will perform better for high end gfx setups. As it is now it feels similar to MS FSX where the CPU definitely was the bottle neck making powerful gfx cards a waste. Even though not as bad.

 

It would be interesting to have some more comparisons with i5 versus i7 CPU’s.

Link to comment
Share on other sites

BitMaster: Note that the ~100fps numbers are average framerate while playing back a specific track. Look at the max fps, and you see that it even approaches 200fps is some cases. Obviously there's no fps cap involved. I'm absolutely certain that if you played back the same track on your PC, you would see no more than around 100fps average either.

 

 

 

yep,, my fault...didnt see the dotted lines beyond 100fps...guess I need glasses.

 

No...180~190fps seems ok...thats where I max out too in FC3 AC or Su25T

 

 

 

but I found another thing that I cant believe now: DCS uses 4 cores !??? That can only come if you run a HT CPU and your core splitts the thread across the 2 HT cores, this wont happen on a Core i5 I guess, but imight be wrong, only ever had i7 or other fully equipped models


Edited by BitMaster

Gigabyte Aorus X570S Master - Ryzen 5900X - Gskill 64GB 3200/CL14@3600/CL14 - Asus 1080ti EK-waterblock - 4x Samsung 980Pro 1TB - 1x Samsung 870 Evo 1TB - 1x SanDisc 120GB SSD - Heatkiller IV - MoRa3-360LT@9x120mm Noctua F12 - Corsair AXi-1200 - TiR5-Pro - Warthog Hotas - Saitek Combat Pedals - Asus PG278Q 27" QHD Gsync 144Hz - Corsair K70 RGB Pro - Win11 Pro/Linux - Phanteks Evolv-X 

Link to comment
Share on other sites

Tiborrr, apart from my nagging..LOL I admire what you do, you seem to have patience and an analytic mind.

 

If you come across a 2600k, latest batch, grab it ! I would like to see how it performs vs. a brand new 500€ i7 that may not clock as high. I am confident that the 2600k at anywhere beyond 4.5 GHz will match the newest i7's unless they equal Hz.

 

I am holding back to renew cause it runs just fine on my rig and I would like to hear:

 

YES..YOU CAN GET MORE , MUCH MORE...not just 10fps for 2000€ down the road.

 

Bit

Gigabyte Aorus X570S Master - Ryzen 5900X - Gskill 64GB 3200/CL14@3600/CL14 - Asus 1080ti EK-waterblock - 4x Samsung 980Pro 1TB - 1x Samsung 870 Evo 1TB - 1x SanDisc 120GB SSD - Heatkiller IV - MoRa3-360LT@9x120mm Noctua F12 - Corsair AXi-1200 - TiR5-Pro - Warthog Hotas - Saitek Combat Pedals - Asus PG278Q 27" QHD Gsync 144Hz - Corsair K70 RGB Pro - Win11 Pro/Linux - Phanteks Evolv-X 

Link to comment
Share on other sites

BitMaster: Note that the ~100fps numbers are average framerate while playing back a specific track. Look at the max fps, and you see that it even approaches 200fps is some cases. Obviously there's no fps cap involved. I'm absolutely certain that if you played back the same track on your PC, you would see no more than around 100fps average either.

 

i would like to replay the trk file. any link ??

Gigabyte Aorus X570S Master - Ryzen 5900X - Gskill 64GB 3200/CL14@3600/CL14 - Asus 1080ti EK-waterblock - 4x Samsung 980Pro 1TB - 1x Samsung 870 Evo 1TB - 1x SanDisc 120GB SSD - Heatkiller IV - MoRa3-360LT@9x120mm Noctua F12 - Corsair AXi-1200 - TiR5-Pro - Warthog Hotas - Saitek Combat Pedals - Asus PG278Q 27" QHD Gsync 144Hz - Corsair K70 RGB Pro - Win11 Pro/Linux - Phanteks Evolv-X 

Link to comment
Share on other sites

Only problem with 4k is that you need a large monitor

I haven't seen any 4k monitors under 27 inches and am I use a 24 ( 23.5) inch monitor 144 hertz refresh rate. 1080p. Even took me a while to adjust from 22 - 24

 

27 inch is pretty much a small TV. Much too big for my liking probslby not good for your eyes staring so close to a much larger screen and I'd need a new desk to fit the larger monitor especially for multiple monitor hookups

 

:/


Edited by Kev2go

 

Build:

 

Windows 10 64 bit Pro

Case/Tower: Corsair Graphite 760tm ,Asus Strix Z790 Motherboard, Intel Core i7 12700k ,Corsair Vengeance LPX DDR4 64gb ram (3600 mhz) , (Asus strix oc edition) Nvidia RTX 3080 12gb , Evga g2 850 watt psu, Hardrives ; Samsung 970 EVo, , Samsung evo 860 pro 1 TB SSD, Samsung evo 850 pro 1TB SSD,  WD 1TB HDD

 

Link to comment
Share on other sites

Only problem with 4k is that you need a large monitor

I haven't seen any 4k monitors under 27 inches and am I use a 24 ( 23.5) inch monitor 144 hertz refresh rate. 1080p. Even took me a while to adjust from 22 - 24

 

27 inch is pretty much a small TV. Much too big for my liking probslby not good for your eyes staring so close to a much larger screen and I'd need a new desk to fit the larger monitor especially for multiple monitor hookups

 

:/

 

4K on anything smaller than 27 inch is wasted pixels. Not to mention all of the problems you get with unreadable text.

Link to comment
Share on other sites

Question for you guys. Of if anyone can bench a n i7 quad and a i7 hex core processor and adjust clocks so they match in an effort to see if cpu utilisation of dcs goes beyond four cores? We know clock speed is king in dcs, im just curious if hex cores make a difference should one were to compare of a quad at the same speed.. thanks!

find me on steam! username: Hannibal_A101A

http://steamcommunity.com/profiles/76561197969447179

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...