Low CPU utilization & Low FPS

[[frl]myke 14]

Also a lot of people don't come to this forum and probably don't realize it exists, and they're having issue's.

True but...

Also a lot of people don't come to this forum and probably don't realize it exists, and they're not having issue's.

Can we agree that we both have the same ArmA 3? Now please, explain me why i have 50-60 FPS in SP. Explain me please. According to you this isn't possible. It might be related to the fact that i don't try to play on ultra but this is a extremely wild speculation. Now please, go ahead, explain my good FPS, i'm really interested to know.

[MavericK96 0]

Myke;2680064']The point is: complaining doesn't help' date=' proper bug reporting does. Just putted it in the right words. It's not rocket science, really.[/quote']

Well, no offense, but there is/was a massive ticket for this issue (Low FPS/Low Utilization) on the Tracker for months, if not years (for ArmA 2 as well) and it has been barely acknowledged.

EDIT: For reference: http://feedback.arma3.com/view.php?id=716

2200+ votes, submitted March 2013, "assigned" but no developer comment since...ever? There may be one in there, but if so I can't find it.

Edited May 1, 2014 by MavericK96

[[frl]myke 14]

Well, no offense, but there is/was a massive ticket for this issue (Low FPS/Low Utilization) on the Tracker for months, if not years (for ArmA 2 as well) and it has been barely acknowledged.

EDIT: For reference: http://feedback.arma3.com/view.php?id=716

2200+ votes, submitted March 2013, "assigned" but no developer comment since...ever? There may be one in there, but if so I can't find it.

Did you actually read through all the posts? People using it more like a forum than a bugtracker. It's unbelievable how people waste dev's time by forcing them to dig through unhelpful comments. Not really motivating.

Personally, i would get a lot more motivation out of a "nice work but there are a few rather serious issues to look into" rather than "it runs like s**t, fix it nao".

[F3r4_N3n3m 10]

The Topic is about LOW CPU AND GPU UTILIZATION. So let's talk about that.

I get GPU utilization of 20% on Ultra and of 5% on Standard/High. Actually the graphic Card goes into idle when I use Standard/High because it believes it is not being used.

The CPU utilization never goes more than 60% in any mode or time. It is however very unstable in all cores all the time. This Kind of behavior I have only in Arma.

WHAT CAN I DO TO INCREASE CPU AND GPU UTILIZATION? Not talking about fps.

[windies 11]

Myke;2680342']Did you actually read through all the posts? People using it more like a forum than a bugtracker. It's unbelievable how people waste dev's time by forcing them to dig through unhelpful comments. Not really motivating.

Personally' date=' i would get a lot more motivation out of a "nice work but there are a few rather serious issues to look into" rather than "it runs like s**t, fix it nao".[/quote']

Again, you stereotype everyone with a problem into the "it runs like s**t, fix it nao" crowd. You act like none of us have tried to repro, tried to offer constructive criticism or even are still providing constructive criticism when asked. Just like you acted like anyone with a problem must be running 5 year old PC's with SUPER DUPER ULTRA settings. We can't help people who act like true tards or idiots from writing out complaints, but you seem to group anyone with a problem with them.

Also, It's their effing job man, you act like they are primadonna's who can't do anything without the right circumstances or need to be patted on the back constantly when they trip up or screw up. I mean seriously, give credit where credit is due but you don't have to kiss ass with every critical comment you make in order to be helpful as long as you're not just trying to bash the crap out of the developers or something.

[oddballsix 19]

I built out a load test mission that is 64 helicopter AI units from 3 sides getting into an air battle over the main airport. Added to which are a handful of ground units and a single plane.

http://www.wanderson.org/arma/Furball.Altis.zip

Running through this custom mission, I started a windows performance log while I played through the mission.

http://www.wanderson.org/arma/ArmaPerformance.blg

Concurrently, I was also logging a graphics card session using the old Riva stuff built into the eVGA PrecisionX tool. I didn't post it, because it doesn't matter, as I will talk about in a minute. My GPU never got even scratched.

Platform:

Windows 8.1 Update

Intel Core i7-930 @ 2.8 GHz

12GB DDR3 Triple-Channel SDRAM @ 1033Mhz

eVGA GeForce 780 SC with ACX Cooler

GeForce 335.28 Driver

System Drive is a Crucial M4 SSD

Application/Data drive for the application is a RevoDrive PCI-E hybrid using the nvelo dataplex software.

Key observations:

• 2+ GB of private memory allocation
  
• There are a lot of pagefile operations.
  
• In a heavy-weight AI mission like the torture test this created mission represents, the actual graphics card load is so low that the graphics card is never spinning up significantly, operating at ~360Mhz instead of the ~1GHz max sustained clock rate it is capable of
  
• The monitored disk loads that are specific to the process actually are not that bad, just as you might expect.
  
• The system activity - which includes page file actions - is where the disk operations actually are and may have a correlation to performance indicators.
  

Edited to add:

Thinking more about the Windows 8 mechanics, I generalized some of my text above. Difficult to draw conclusions and I am suspicious because the memory allocation maps exactly, byte for byte, with claimed process-associated page file utilization. Indicates that windows' perfmon capture may be inaccurate for determining the cause of the actual page file utilization.

Edited May 2, 2014 by OddballSix

[[frl]myke 14]

Again, you stereotype everyone with a problem into the "it runs like s**t, fix it nao" crowd. You act like none of us have tried to repro, tried to offer constructive criticism or even are still providing constructive criticism when asked. Just like you acted like anyone with a problem must be running 5 year old PC's with SUPER DUPER ULTRA settings. We can't help people who act like true tards or idiots from writing out complaints, but you seem to group anyone with a problem with them.

Also, It's their effing job man, you act like they are primadonna's who can't do anything without the right circumstances or need to be patted on the back constantly when they trip up or screw up. I mean seriously, give credit where credit is due but you don't have to kiss ass with every critical comment you make in order to be helpful as long as you're not just trying to bash the crap out of the developers or something.

I'm sorry, i'm only responsible for what i write, not for what you understand. Read through the bug feedback Maverick linked and see how many useless posts are in there. And please please, tell me where i said "all" or "everyone". I just said that those "fix it nao" peoples most likely get ignored, simply because their "feedback" doesn't contain any valuable info.

And since you're accusing me of stereotyping, could you please answer my question from here: http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2680317&viewfull=1#post2680317

Thank you. Please stop picking arguments just to prove your point. either answer all or nothing.

---------- Post added at 00:47 ---------- Previous post was at 00:43 ----------

The Topic is about LOW CPU AND GPU UTILIZATION. So let's talk about that.

I get GPU utilization of 20% on Ultra and of 5% on Standard/High. Actually the graphic Card goes into idle when I use Standard/High because it believes it is not being used.

The CPU utilization never goes more than 60% in any mode or time. It is however very unstable in all cores all the time. This Kind of behavior I have only in Arma.

WHAT CAN I DO TO INCREASE CPU AND GPU UTILIZATION? Not talking about fps.

NOTHING....erm, sorry...nothing. CPU load represents the usual downsides of multithreading: threads waiting for data from other threads, therefor idling. GPU waits for data from the CPU to render the scene. If CPU is under heavy load (which isn't represented by % number), the GPU load goes down. You might want to invest the idling power of the GPU to raise GPU related graphic settings.

[Varanon 892]

Myke;2680342']Personally' date=' i would get a lot more motivation out of a "nice work but there are a few rather serious issues to look into" rather than "it runs like s**t, fix it nao".[/quote']

No offense, but I do this sort of bug reporting every time (honestly, I try to be polite and assertive). A lot of good feedback has been given on these issues, especially the performance issue, so would you please not make us look like whiners and ragers that shout "fix this shit", because we sure as hell don't do that ?

[windies 11]

Myke;2680368']I'm sorry' date=' i'm only responsible for what i write, not for what you understand. Read through the bug feedback Maverick linked and see how many useless posts are in there. And please please, tell me where i said "all" or "everyone". I just said that those "fix it nao" peoples most likely get ignored, simply because their "feedback" doesn't contain any valuable info.

And since you're accusing me of stereotyping, could you please answer my question from here: [url']http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2680317&viewfull=1#post2680317[/url]

Thank you. Please stop picking arguments just to prove your point. either answer all or nothing.

---------- Post added at 00:47 ---------- Previous post was at 00:43 ----------

NOTHING....erm, sorry...nothing. CPU load represents the usual downsides of multithreading: threads waiting for data from other threads, therefor idling. GPU waits for data from the CPU to render the scene. If CPU is under heavy load (which isn't represented by % number), the GPU load goes down. You might want to invest the idling power of the GPU to raise GPU related graphic settings.

I'm not the only one misunderstanding you then apparently but nice try at deflecting this on to me. You're the one making the claims and telling people they're wrong and stereotyping them whether you yourself understand that or not. And I quote:

No offense, but I do this sort of bug reporting every time (honestly, I try to be polite and assertive). A lot of good feedback has been given on these issues, especially the performance issue, so would you please not make us look like whiners and ragers that shout "fix this shit", because we sure as hell don't do that ?

So it isn't just me that thinks you're being a bit stereotypical and narrow in your points of view and responses. But hey, keep thinking the world revolves around you and this forum, it's working for ya.

As for a answer to your question, which has nothing to do with you stereotyping so I have no idea what you're on about there, give me some actual specifics like what part of the map, whether it's SP or MP, what actual settings you are using etc..., otherwise honestly I think you're full of crap if you're telling me you get 50-60 fps in MP with any actual kind of mission with moderate load.

Edited May 1, 2014 by Windies

[Th4d 10]

No offense, but I do this sort of bug reporting every time (honestly, I try to be polite and assertive). A lot of good feedback has been given on these issues, especially the performance issue, so would you please not make us look like whiners and ragers that shout "fix this shit", because we sure as hell don't do that ?

Fanboys can´t help themselves, they ignore everything shoved in their faces and dismiss it as a minority even when people on this forum make a poll and turns out 50% of around 1800 are displeased with the multiplayer performance. That´s the official stance, this developer cannot handle the issue, the only thing left to do is make empty promises that are around for years and belittle anyone who complains.

The diagnosis for the mental disorder on such fanboys is cognitive dissonance, because i truly think that he does believe in what he says.

http://en.wikipedia.org/wiki/Cognitive_dissonance

Edited May 1, 2014 by Th4d

[[frl]myke 14]

otherwise honestly I think you're full of crap if you're telling me you get 50-60 fps in MP

Thank you, thank you so much. Best proof that you just read what you want to read. Let me quote myself:

Now please, explain me why i have 50-60 FPS in SP.

Again, thank you, thank you. :icon_eek:

---------- Post added at 01:22 ---------- Previous post was at 01:20 ----------

No offense, but I do this sort of bug reporting every time (honestly, I try to be polite and assertive). A lot of good feedback has been given on these issues, especially the performance issue, so would you please not make us look like whiners and ragers that shout "fix this shit", because we sure as hell don't do that ?

True that, wasn't my intention to forget all those really helpful people who actually give really good bug reports. My apologies.

Edited May 1, 2014 by [FRL]Myke

[windies 11]

Myke;2680405']Thank you' date=' thank you so much. Best proof that you just read what you want to read. Let me quote myself:

Again, thank you, thank you. :icon_eek:

---------- Post added at 01:22 ---------- Previous post was at 01:20 ----------

What are you doing talking about SP when we're all talking about MP then? This is what I'm talking about when I say you're biased and ignorant. I miss 2 letters in your post but you miss the entire point of the conversation or problems half the time. Seriously, I glossed over 2 letters in your post, whoops my bad. It's hard not to when the rest of your posts are such atrocious self absorbed diatribes towards people you dislike because they don't share the same biased outlook as you.

True that, wasn't my intention to forget all those really helpful people who actually give really good bug reports. My apologies.

Very subtle... :rolleyes: Guess I shouldn't expect any more from a biased moderator.

Edited May 2, 2014 by Windies

[MavericK96 0]

Myke;2680342']Did you actually read through all the posts? People using it more like a forum than a bugtracker. It's unbelievable how people waste dev's time by forcing them to dig through unhelpful comments. Not really motivating.

Personally' date=' i would get a lot more motivation out of a "nice work but there are a few rather serious issues to look into" rather than "it runs like s**t, fix it nao".[/quote']

So what the developers need to do is make a STANDARD benchmark, that everyone can run and will be EXACTLY the same, and then ask for specific details on performance.

The issue has been known for a long time. I don't know what more you expect the community to do. Several people have posted their frame rate results in this thread, but it's with varying missions so that probably isn't that helpful. I have seen dozens of posts suggesting a standardized benchmark/performance routine to post that the developers design and will take seriously, and as of yet I have not seen this come to light. It gives a feeling that this is not a priority, when the same issue has been prevalent since ArmA 2 and possibly even before.

As for "motivation"...you can choose what comments you take seriously, and which you disregard as trolling/spam. I agree that there are a lot of unhelpful comments, but does that really excuse the fact that there has been no real progress made on that ticket, at least that any of us have had any visibility on?

I mean...I get what you are saying but it sounds as if you are excusing a complete lack of progress/visibility with, "well, people write unhelpful comments and it makes the devs cry, so they don't work on it"...

Really? I think they are adults.

[windies 11]

So what the developers need to do is make a STANDARD benchmark, that everyone can run and will be EXACTLY the same, and then ask for specific details on performance.

The issue has been known for a long time. I don't know what more you expect the community to do. Several people have posted their frame rate results in this thread, but it's with varying missions so that probably isn't that helpful. I have seen dozens of posts suggesting a standardized benchmark/performance routine to post that the developers design and will take seriously, and as of yet I have not seen this come to light. It gives a feeling that this is not a priority, when the same issue has been prevalent since ArmA 2 and possibly even before.

As for "motivation"...you can choose what comments you take seriously, and which you disregard as trolling/spam. I agree that there are a lot of unhelpful comments, but does that really excuse the fact that there has been no real progress made on that ticket, at least that any of us have had any visibility on?

I mean...I get what you are saying but it sounds as if you are excusing a complete lack of progress/visibility with, "well, people write unhelpful comments and it makes the devs cry, so they don't work on it"...

Really? I think they are adults.

This, exactly this.

[oddballsix 19]

What are you doing talking about SP when we're all talking about MP then? This is what I'm talking about when I say you're biased and ignorant. I miss 2 letters in your post but you miss the entire point of the conversation or problems half the time. Seriously, I glossed over 2 letters in your post, whoops my bad. It's hard not to when the rest of your posts are such atrocious self absorbed diatribes towards people you dislike because they don't share the same biased outlook as you.

The level of vitriol and personal attacks in this thread are getting excessive.

1) Putting any bias aside, people using the bug tracker to rant or as a forum back and forth make it harder for the developers to extract useful information form the bug system. It also decreases the emphasis they may put on user generated concerns given how much 'chaff' they get in the system versus 'wheat' that allows them to move forward.

2) Performance is an issue in both MP *and* SP. This thread is (and probably should be) about both. As I posted many, many pages ago in this thread, Multiplayer performance is very challenging to troubleshoot. SP performance allows systemic performance characteristics to be isolated away from the netcode and server interaction. There is then a second layer of performance elements that need to be addressed in the server/client interaction, together with compensation for internet latency.

3) The claim that the developers are not aware of the performance issues, don't care, haven't responded, etc is simply patently false. Not only have they confirmed they read this thread almost every day, interacted, asked questions, and engaged in other community threads on the subject, the roadmap published today squarely addresses this issue.

However, while it's great to have one eye on the future, we'd also like to stress that our original (and ongoing) goals haven't changed. It (almost) goes without saying that our work on stability and optimization remains central to Arma 3's longevity. Working closely with community groups to test large-scale MP performance continues, as does our cooperation with the DayZ team and their knowledge of MP optimization techniques. That aside, our own official multiplayer servers provide a wealth of useful data, which feeds back into development.

[MavericK96 0]

And that's great...I am glad they addressed it officially, though from reading the whole roadmap I can't help but feel like performance is secondary (or even tertiary) to DLC and mission content. Don't get me wrong, the new features and DLC sound really awesome (firing from vehicles?! Finally!) but personally I would prefer to see performance in the spotlight as opposed to a small snippet at the end of a considerable report. It's just an issue that affects the entire game and the enjoyment of it. No amount of new content is going to make the game enjoyable if you are getting 15-20 FPS (which regularly happens for me on what I think is still a quite capable PC) in both SP and MP.

Still, I am appreciative that performance was brought up in the roadmap. I hope we see a closer effort between the developers and community to test and fix the issues.

[F3r4_N3n3m 10]

Myke;2680368']NOTHING....erm' date=' sorry...nothing. CPU load represents the usual downsides of multithreading: threads waiting for data from other threads, therefor idling. GPU waits for data from the CPU to render the scene. If CPU is under heavy load (which isn't represented by % number), the GPU load goes down. You might want to invest the idling power of the GPU to raise GPU related graphic settings.[/quote']

Ok, so I understand that the CPU load is keeping my GPU from working as it should.

But my CPU load is not Close to being high. Does that mean I have a Problem with my System? Can't I optmize this somehow?

If the CPU utilization is the bottleneck for Performance in this application, shouldn't I be experiencing good Performance when the System is under utilized?

[[frl]myke 14]

Ok, so I understand that the CPU load is keeping my GPU from working as it should.

But my CPU load is not Close to being high. Does that mean I have a Problem with my System? Can't I optmize this somehow?

If the CPU utilization is the bottleneck for Performance in this application, shouldn't I be experiencing good Performance when the System is under utilized?

No, you don't have a problem with your system. Look at it this way: there are 4 workers but three of them regulary have to wait for data from the fourth to continue so they wait, idling, and you see 25% workload.

There are many things that rely on each other: AI, Physics, rendering, collision and probably a lot of other tasks . The fact that ArmA 3 isn't a player-centric game like almost every other shooter actually available doesn't help to make things easier.

So, a faster CPU can help since it get's more done due to more cycles per second but in task manager you'll probably see the same workload (25%, taking the example above) as with a slower CPU.

And that's also why it isn't that easy to improve performance. If you change one thing you have to make sure the others don't get broken because of the change you've made (see also my signature). This is probably the case when people report lower FPS after a patch. They (probably) optimized one part which gives 3FPS more but another part gets negatively influenced and drops then by 5FPS.

Optimizing therefor is extremely difficult to not break the whole game, that's why it takes long.

[nikiforos 450]

I personally have accepted that we are not going to see any optimization now or in the future. I would be a fool to believe in something that I have expected for many many years but never seen anything good from these discussions.

I'm not saying that BIS totally ignore this issue but I believe that they know that they can't do much to improve the situation. The best thing to do if you know that you can handle a situation is to ignore it and to stop involving in debates and discussions. And I think that this is the reason we never saw the optimizations blog they talked about in the past.

Edited May 2, 2014 by Nikiforos

[LSD_Timewarp82 21]

This is so ridicolous that the newest CPU Technology (my second pc with the 3770K @ 4,5Ghz) with the CPU gimmicks should not be able to work correctly with Arma3 like Myke said "AI, Physics, rendering, collision and probably a lot of other tasks"

I think the Arma Engine maybe has those things deactivated, thats why even the fastest CPU is running at these low workload every user see.

[[frl]myke 14]

This is so ridicolous that the newest CPU Technology (my second pc with the 3770K @ 4,5Ghz) with the CPU gimmicks should not be able to work correctly with Arma3 like Myke said "AI, Physics, rendering, collision and probably a lot of other tasks"

I think the Arma Engine maybe has those things deactivated, thats why even the fastest CPU is running at these low workload every user see.

There are many things that rely on each other

If everything is coupled and one thread has to wait for another thread because it needs that data to continue......is it really that hard to understand? Gimmicks don't help. You have 4 workers, 3 of them are idle because they need to wait for the fourth to finish, 25% workload, 75% idle. Graphics renderer can't draw the bullet before physics determined it's exact location and heading at a given time. Just to name one example out of thousands that happen every frame.

[St. Jimmy 272]

Myke;2680753']If everything is coupled and one thread has to wait for another thread because it needs that data to continue......is it really that hard to understand? Gimmicks don't help. You have 4 workers' date=' 3 of them are idle because they need to wait for the fourth to finish, 25% workload, 75% idle. Graphics renderer can't draw the bullet before physics determined it's exact location and heading at a given time. Just to name one example out of thousands that happen every frame.[/quote']

Would it be ever possible so we render things faster that the simulation goes? So that simulation happens with 20fps but the scene is drawed with 60fps. Maybe already happens that way, I'm not a programmer so I don't know, but I wonder then if they could squeeze even more.

At least DayZ is doing something like this:

Graphics Optimization and the Renderer

Work has commenced on disconnecting the renderer directly from the simulation. This would then free us to change (or even replace) the renderer much easier, allowing greater optimization and new functionality such as better particle effects. We have no timeframe for this completion no any specific goals beyond preparing it for future goals, but it’s exciting for us to start this work.

This work will offer us the potential for future very significant client FPS optimizations, such as better scene management, to solve issues people experience such as low FPS in cities.

Sounds like a big step. But I wonder is the new server-client architecture that allows to do that or could that be done in Arma also.

[jumpinghubert 48]

So what the developers need to do is make a STANDARD benchmark, that everyone can run and will be EXACTLY the same, and then ask for specific details on performance.

please everyone here put this statement into your signature. Because a mp and sp benchmark is the most helpful thing and the next logical step to adress the problem in a qualitative AND quantitative way and helps the developers too to check every devbranch performancewise.

[Peter5on 10]

I have one core around 90% and the other three are around 50% almost all time. Gpu is around 40-70% So i think its not going to be better, only thing we could do is to wait for Broadwell processors.

So i think Devs should change Arma 3 system requirements to:

-Only Intel i7 4770K and OC to 4.5ghz otherwise stay away.

-U have AMD processor... WRONG stay away.

Is Arma going to be 64bit like Day Z?

[nikiforos 450]

please everyone here put this statement into your signature. Because a mp and sp benchmark is the most helpful thing and the next logical step to adress the problem in a qualitative AND quantitative way and helps the developers too to check every devbranch performancewise.

Well I don't think a standard benchmark is in BIS interests. Just think about it.