DCS 2.0 Nevada stutter help

Man, I feel blessed. I haven't experience any stuttering/freezing. While I own just about every module, I typically only fly the Huey, A10, and Mustang. I could be mistaken but I think more people are having problems with Sukhoi lines.

Hi All,

 

I know this has probably been asked before so apologies but im getting a few stutters in dcs 2.0 nevada usaully when turning, dont have mirrors on, using a GTX 980Ti card, wondered if there any performance tips i could try

 

 

PC Specs in my signature

 

Thanks

 

Gazmon

Switch from an HDD to an SSD and your stuttering will be solved. A PC with a good card, cpu, and RAM will be severely bottlenecked by an HDD.

The stuttering is not your computer ! When you have 16gb of ram and your running a gtx980Ti together with a powerfull CPU its clearly something in the code of the eagle dynamics graphics engine thats not functioning properly.

 

I have these stutters aswell and when i track my fps it stays at 60 constantly and still the game has like these mini freezes and stutters.

 

But hey its in Alpha stage so eventually all the stutters and buggers will be worked out :)

+1

Switch from an HDD to an SSD and your stuttering will be solved. A PC with a good card, cpu, and RAM will be severely bottlenecked by an HDD.

Not true unfortunately. I have an SSD and still have stuttering. It's an engine problem and not a hardware problem.

Not true unfortunately. I have an SSD and still have stuttering. It's an engine problem and not a hardware problem.

When the stuttering starts alt tab out to windows and make sure there is no active process using core1...close it if there is.

Monitor your gpu temperature while playing..it could be on the high side and your gpu clock is stepping down because of it.Id recommend using a program that lets you set your fan speed a bit higher while playing.

Man, I feel blessed. I haven't experience any stuttering/freezing. While I own just about every module, I typically only fly the Huey, A10, and Mustang. I could be mistaken but I think more people are having problems with Sukhoi lines.

Mate, you have a Titan X, surely you are blessed. :smilewink:

Not true unfortunately. I have an SSD and still have stuttering. It's an engine problem and not a hardware problem.

I agree too.

If that is true why do I not get any stuttering?

Probably because you are only running 1080p? I get no stuttering at that res to but once I enable 3D or run 4K then Vegas and bigger cities stutter

This may help: https://www.naturalpoint.com/trackir/downloads/

"Fixed an issue where certain NVIDIA GPUs might step down to a lower power state while the TrackIR software was active."

Not true unfortunately. I have an SSD and still have stuttering. It's an engine problem and not a hardware problem.

Bummer, man. I have a 4790K, 970 ftw, and 16 ram...but was running DCS on an HDD and the stuttering was bad (as were the loading times).

I purchased a Samsung 850 pro SSD and that solved the problem for me. How fast is your SSD?

This may help: https://www.naturalpoint.com/trackir/downloads/

 

"Fixed an issue where certain NVIDIA GPUs might step down to a lower power state while the TrackIR software was active."

That's an interesting one, will check this out

Sent from my iPhone using Tapatalk

Updating trackir worked wonders for me! Thanks!

I have severe stuttering here too, so much it makes it very difficult to play campaigns like red flag for the A-10C. My specs are below in signature, and the settings are in the screenshot (I think it's relevant to give at least this information if we want to make any comparison).

I was first on "Visib range" ultra, but had to give up even though on the Caucasian map it's perfectly fine.

What I see is a lot of swapping, I get on average 20-30k page faults per second - that's when it's not too bad, and constant I/O reads but I'm not sure those are only from the paging. I set a constant paging file at Windows installation (I'm usually doing that to avoid fragmentation, but this time it's on a SSD anyway). The OS is on SSD, but not DCS.

What is strange is that the virtual space taken by the program is about 9-10 GB so it should fit in RAM. I'm wondering whether those I/O are coming from constant texture / terrain loading, which is a known problem for example on P3D with high-resolution textures and meshes. But the swapping needs to be resolved first.

I don't believe TrackIR has anything to do with that (in my case anyway), nevertheless I updated it to 5.4.1 just to be safe. If you do that, don't forget to save your settings because they still haven't understood the principle of user settings, and stubbornly store them in the program area!

I'll compare that with the main branch when I have some time.

Also the loading time (the first time) of NTTR is very, very, very long.

heat blur off

cockpit res 1024 every fram

may be shadows if your changing view s

I have severe stuttering here too, so much it makes it very difficult to play campaigns like red flag for the A-10C. My specs are below in signature, and the settings are in the screenshot (I think it's relevant to give at least this information if we want to make any comparison).

 

I was first on "Visib range" ultra, but had to give up even though on the Caucasian map it's perfectly fine.

 

What I see is a lot of swapping, I get on average 20-30k page faults per second - that's when it's not too bad, and constant I/O reads but I'm not sure those are only from the paging. I set a constant paging file at Windows installation (I'm usually doing that to avoid fragmentation, but this time it's on a SSD anyway). The OS is on SSD, but not DCS.

 

What is strange is that the virtual space taken by the program is about 9-10 GB so it should fit in RAM. I'm wondering whether those I/O are coming from constant texture / terrain loading, which is a known problem for example on P3D with high-resolution textures and meshes. But the swapping needs to be resolved first.

 

I don't believe TrackIR has anything to do with that (in my case anyway), nevertheless I updated it to 5.4.1 just to be safe. If you do that, don't forget to save your settings because they still haven't understood the principle of user settings, and stubbornly store them in the program area!

 

I'll compare that with the main branch when I have some time.

 

Also the loading time (the first time) of NTTR is very, very, very long.

The Red flag campaigns act like stuttermachines on my computer.

Even so much that i just have given up for now...

Im running DCS on a 3770K Oc'ed 4.8Ghz watercooled, 980Ti, 16gb Ram, 2560x1440 resolution. My OS is on an SSD and DCS is installed on 2 spinpoint harddrives in raid 0 config.

The point is flying solo i have 0 troubles or stutters or whatever. As soon as airtraffic is inserted in mass like in the red flag campaigns the stuttering and lagging begins.

My wife was so kind to buy me a Samsung 850Evo 500Gb SSD. This will probably arrive somewhere on wednesday. Then i will reinstall DCS on the SSD and see if this makes a difference.

Im also gonna do a little test in the editor with waypoints and without waypoints. Because i have a strong feeling the scripts that are being used for the campaigns are causing alot of trouble.

I'm experiencing similar problems in openbeta too: missions with few IA interactions work like a charm, when massive IA comes to play (200+ air/land units ) then the sim stutters a lot even if looking an empty little portion of the map

I have two Ssd one for os and one for Dcs, I still get mad stutters in the red flag campaign in 2.0. Empty map no problem 476 Nevada mission or the a10c red flag and fps and stutter problems.

Looks like there is a serious issue with 2.0 stuttering for lots of people, maybe be nice to see some ED forum input, or developer feedback here now.

Maybe needs escalating, think we might be past theories probables etc

Sent from my iPad using Tapatalk

It's beta and alpha I'm sure it will get worked out but in mean time if devs need logs or what not please let us know.

I posted logs that never seem to get me any feedback?? Some problems I don't think even ED can answer there selves?

Sent from my iPad using Tapatalk

As I said, I think we need to compare real data to reach any conclusion.

In the mean time, I did a quick comparison of the A-10C Red Flag campaign on 2.0.1.49921.15 and the A-10C Georgian campaign on 1.5.2.49392.141. But I also get those page faults at similar rates in the 1.5 version, with no apparent stuttering. The only significant difference is the virtual memory size.

The settings are almost the same, in 1.5 I left visibility range on Ultra (and the cockpit resolution is higher, I don't think this is relevant here) and still it doesn't stutter in a visible way. Arguably there are fewer details on the map.

In both cases the physical memory is much lower than the virtual memory, 1 GB more taken by 2.0 than 1.5. And there is still 9 GB of free physical memory, only about half the pages are physical (with 2.0).

Another comparison between 2.0 visibility range medium, high and ultra shows the total virtual memory taken does not change much. And indeed, I still get massive stuttering on medium; I suppose the texture quality would have even less influence (on a high-end graphics card anyway).

Is the terrain completely pre-loaded? The pre-load radius is max'ed at 150000 m. In the manual, it's said that a large value may reduce swapping provided there is enough RAM. Obviously it's not working well enough. I haven't verified and lowered this value though, but it doesn't look like a possible solution.

Is it more efficient to pre-load everything and let the OS manage the virtual memory, or lock the virtual pages in physical memory to swap them while they are not visible and will not cause stuttering?

Also, normally the OS manages the working set on a first-come, first-serve basis, so if no other process is claiming memory, those loaded pages should remain in physical memory. Unless this has changed, I don't know enough about Windows 7 to confirm that.

Attachments: 1.5.jpg = DCS.exe in 1.5, 2.0.jpg is in 2.0, and the 2.0_s.2.jpg is the system load in 2.0.

DCS 2.0 Nevada stutter help

You know there would be a really good starting point here if......ED devs working on 2.0 posted there PC specs or test systems

As I'm sure they don't get any stutters

Sent from my iPad using Tapatalk

As I said, I think we need to compare real data to reach any conclusion.

 

In the mean time, I did a quick comparison of the A-10C Red Flag campaign on 2.0.1.49921.15 and the A-10C Georgian campaign on 1.5.2.49392.141. But I also get those page faults at similar rates in the 1.5 version, with no apparent stuttering. The only significant difference is the virtual memory size.

 

The settings are almost the same, in 1.5 I left visibility range on Ultra (and the cockpit resolution is higher, I don't think this is relevant here) and still it doesn't stutter in a visible way. Arguably there are fewer details on the map.

 

In both cases the physical memory is much lower than the virtual memory, 1 GB more taken by 2.0 than 1.5. And there is still 9 GB of free physical memory, only about half the pages are physical (with 2.0).

 

Another comparison between 2.0 visibility range medium, high and ultra shows the total virtual memory taken does not change much. And indeed, I still get massive stuttering on medium; I suppose the texture quality would have even less influence (on a high-end graphics card anyway).

 

Is the terrain completely pre-loaded? The pre-load radius is max'ed at 150000 m. In the manual, it's said that a large value may reduce swapping provided there is enough RAM. Obviously it's not working well enough. I haven't verified and lowered this value though, but it doesn't look like a possible solution.

 

Is it more efficient to pre-load everything and let the OS manage the virtual memory, or lock the virtual pages in physical memory to swap them while they are not visible and will not cause stuttering?

 

Also, normally the OS manages the working set on a first-come, first-serve basis, so if no other process is claiming memory, those loaded pages should remain in physical memory. Unless this has changed, I don't know enough about Windows 7 to confirm that.

 

Attachments: 1.5.jpg = DCS.exe in 1.5, 2.0.jpg is in 2.0, and the 2.0_s.2.jpg is the system load in 2.0.

We have a very similar system be interesting to see if you change those settings and it produces a positive result. Keep up good work! Thanks for the info

You know there would be a really good starting point here if......ED devs working on 2.0 posted there PC specs or test systems

 

As I'm sure they don't get any stutters

 

 

Sent from my iPad using Tapatalk

I doubt their systems are much different than ours, look at wags in his sig also some of the testers have info in their sigs.