Jump to content
Sign in to follow this  
dasa

CPU VS RAM Performance & CPU Threading Benchmarked

Recommended Posts

Started out as comparing cpu threads performance for interest sake but it turned into something more

19783c1b-6996-44c8-8e90-8206dea4de70_zps1f1b91c2.jpg

(Edit: Just tested 4.7GHz DDR1333 7-7-7-24-1T 38.9FPS)

Test System

2600k

ASUS Z77 MAXIMUS V GENE

2X4G SAMSUNG 2133 (9-10-10-24-1T used for all tests)

Two 7970 (crossfire disabled) 1185\7000MHz

2TB 7200RPM with OCZ Synapse 128G Cache

Seasonic 1000W Platinum

win7 64

cat 13.3 beta3

Tests were run at maximum detail 720x480 aa disabled but fxxaa on

Some may think running such a low res is stupid but it reduces the effect the gpu has on the tests so that the results are not just margin of error and if i had a gpu capable of running 1920x1080 close to 60fps i suspect it would give these same cpu bottelnecked numbers anyway maybe if we get some drivers that make crossfire work better i will test again at a higher res

i used this benchmark to do the tests http://forums.bistudio.com/showthread.php?151794-ArmA3Mark-Benchmark-your-ArmA-3

performance gain from 1333-2133 ram when at the same timings goes from ~11% at 3.8ghz to ~14% at 4.7ghz but 1600 to 2133 is only ~8%

running different timing may have been interesting aswell but that shall do for now

here is the other test with core scaling the last three tests are just margin of error difference

arma3alphacorescaling_zps4d63d502.png

Edited by dasa

Share this post


Link to post
Share on other sites

was -exThreads=7 used?

Edited by Dwarden

Share this post


Link to post
Share on other sites
was exThreads=7 used?

tried those commands they didnt seem to make any improvement although they could be used to limit the number of cores used

arma was spreading the load across the cores but the total load was only ever around ~90% for two cores or ~40% over 4 cores

like most game engines its threading beyond two cores just isnt that good

Share this post


Link to post
Share on other sites

nah the question was more related to the if no other value was used,

for properly detected quadcore, the default is 7

while for dualcore the default is 3

Share this post


Link to post
Share on other sites
nah the question was more related to the if no other value was used,

for properly detected quadcore, the default is 7

while for dualcore the default is 3

played with these in launch options at first

-cpuCount=4 -exThreads=7

but in the end the only command i used was -noSplash for the tests i linked above and just used task manager to limit the threads arma could use

or is there a .ini\cfg file i can check to see what the game engine has detected?

Share this post


Link to post
Share on other sites

wow, interresting results! And i always thought, clocking ram, won't bring anything at all.

Share this post


Link to post
Share on other sites
was -exThreads=7 used?

really, you are implying that more than 2 cores makes a difference? god, thanks for ignoring every post proving otherwise, good to know all the effort to try to clear the subject from every angle was completely in vain, showing how you guys are indeed not going to do absolutely nothing about it. another half a decade with a dual core game here we go.

il sum it up AGAIN for you: (even though you avoided any previous inquiry i made directly to you in the past in this forum)

http://gamegpu.ru/images/stories/Test_GPU/Action/ARMA%20III%20Alpha/test/arma%203%20proz%20amd.jpg

http://cdn.overclock.net/d/d3/d3796154_proz20amd.jpeg

I started talking about it prior to march 8 in which i proved it here: http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2321697&viewfull=1#post2321697

And the memory frequency comparison posted apply to every software, but tightening timings might wield better results in synthetic benchmarks. (superpi for instance)

Edited by white

Share this post


Link to post
Share on other sites
really, you are implying that more than 2 cores makes a difference? god, thanks for ignoring every post proving otherwise, good to know all the effort to try to clear the subject from every angle was completely in vain, showing how you guys are indeed not going to do absolutely nothing about it. another half a decade with a dual core game here we go.

I simply asked, any reason to troll the topic?

really do You even know what that setting is for ? http://community.bistudio.com/wiki/Arma2:_Startup_Parameters#exThreads

Share this post


Link to post
Share on other sites
really, you are implying that more than 2 cores makes a difference? god, thanks for ignoring every post proving otherwise, good to know all the effort to try to clear the subject from every angle was completely in vain, showing how you guys are indeed not going to do absolutely nothing about it. another half a decade with a dual core game here we go.
he didnt imply anything. He just asked a question.Just another ignore.

Share this post


Link to post
Share on other sites
really, you are implying that more than 2 cores makes a difference? god, thanks for ignoring every post proving otherwise

Not only post, it's the most voted bug in their bug-tracker right now. I won't be surprised if a dev will come up (like it happened last time in ArmA2) to say: "sorry we won't invest time to fix this".. and close the bug... and see you to ArmA4 (probably still running on 2 cores as well..).

Share this post


Link to post
Share on other sites
Started out as comparing cpu threads performance for interest sake but it turned into something more..
Did you turn "core parking" off?

Share this post


Link to post
Share on other sites
I simply asked, any reason to troll the topic?

really do You even know what that setting is for ? http://community.bistudio.com/wiki/Arma2:_Startup_Parameters#exThreads

yes i know what its for, and i used it on the tests i made to prove my point on march 8.

ive used all launch options a ton of times to no avail whatsoever performancewise.

gj trying an ad hominem fallacy (or you simply dont know what trolling is) when you know you are wrong. this subject has been discussed extensively for over a 1 month now and you post that like you are completely oblivious to it.

which means either you are oblivious to it or you just dont care.

and this ia a CPU performance topic. so yes im dead-on on the subject at hand.

---------- Post added at 17:12 ---------- Previous post was at 16:54 ----------

edit after warn:

Dwarden,

first, you dont know what trolling is. but ill explain to you, trolling is when someone wants to make someone else angry by posting a false argument that he knows is false to begin with. so you stating that im trolling makes absolutely no sense whatsoever.

second, his results were OBVIOUS to anyone following the discussing about low CPU performance on you know, that humongous topic about it on throubleshooting, you should be aware of it, you posted on it a long time ago. thus a question about exthreads on those results could only come from someone that didnt read/care about all the material showing that exact same behaviour, in this same forum. which is frustrating to anyone posting on that topic.

which also means i not only was not trolling, but i spoke about CPU performance, questioning a DEV knowledge about it on a CPU performance topic. again, im dead-on on this topics subject.

i understand how frustrated you are of being wrong and want to lash out at me for pointing it out, but you being angry doesnt mean im wrong. use logic and try to prove me wrong then. wait you cant, so yeah the only thing you can do is a forum warn.

very professional. (saving all this as pdf/screenshots in case its erased so i can post on other arma 3 related foruns if necessary.)

Edited by white

Share this post


Link to post
Share on other sites

@DWarden - take no notice.

White, why make yourself sound like a know it all? You could have had a real conversation with a Dev member and actually had some influence for the better, instead you came off this below.

Date joined Mar7 2013

yes i know what its for, and i used it on the tests i made to prove my point on march 8

1 Day spent here, 1 day proving definitive points. . . believes you know engine better than development team.

Edited by Bigpickle

Share this post


Link to post
Share on other sites
@DWarden - take no notice.

Date joined Mar7 2013

1 Day spent here, 1 day proving definitive points. . . believes he knows engine better than development team. Say no more.

i played arma 2 for a long time, never posted before. arma 2 has the exact same issue, always have. and i clearly noticed it again in the first minutes i played arma 3, because this was the one thing i expected to be fixed.

you fail for lack of information and logic.

Share this post


Link to post
Share on other sites

You fail cos I edited my post.

Grow up mate, cos you sound like the typical teenager, I'm done with this now because clearly you are a "wont be told" kinda guy.

Share this post


Link to post
Share on other sites
You fail cos I edited my post.

Grow up mate, cos you sound like the typical teenager.

a lot of people have been asking question directly and politely for a long time, more than a month for arma 3 (years when you add arma 2 to it) and have been either ignored or replied with false vague statements like "no game has high usage across all cores". which i showed easily again as being false in a link in this same topic. so yeah when devs state something that is wrong they either dont know better or are lying.

so its expected for people to get frustrated and be more direct.

and i see no point in partaking on circlejerking or fanboyism. i prefer trying to get the truth, which they avoid at all costs.

i think its funny because you start out with a tipical illogical https://en.wikipedia.org/wiki/Ad_hominem and when you know you have no point but to try to flame me into submission you give up accordingly. which i think its fine.

and i agree answering you have made this thread go way too far from the subject, which is CPU performance and what i was arguing about.

Edited by white

Share this post


Link to post
Share on other sites

I'm not sure, why there's people asking to try this and that, like it is a user "mistake/error", when even the developers recognized that the engine is bugged? We should really read only (from DEV): we promise we'll fix the bug; or (option b): we have no idea how to fix it, because the engine is old and it cannot be fixed. Is that hard?

Share this post


Link to post
Share on other sites
I'm not sure, why there's people asking to try this and that, like it is a user "mistake/error", when even the developers recognized that the engine is bugged? We should really read only (from DEV): we promise we'll fix the bug; or (option b): we have no idea how to fix it, because the engine is old and it cannot be fixed. Is that hard?

Yeah I gotta agree with this. It's very easy for an enthusiast today who overclocks their rig and fiddles around to squeeze performance out of their components to run tests and log data which clearly indicates this game is severely limited in it's ability to use CPU cores. There's an almost infinite amount of reports and data out there on the internet to prove this, yet I have not heard an official response to these cries for improvements. At this point with no proper response to the issue, what can the players think? We think we are being ignored and it's hard to argue that fact when a user makes a legitimate claim (white) and has put in a ton of work demonstrating how this game has not improved upon Arma 2's lack of multithreading, and he gets called a troll... it just blows my mind.

Share this post


Link to post
Share on other sites

... spend some moment on trying find my answers about the performance issues in multiple threads

then you become aware we know about it ...

just go on thread list (not search, not new posts) and there is this little icon saying "Developer posts in thread") click it

and then next to thread # there is icon which takes you to another developer post and so on ...

hard to try it explain to the very same people who ignored the posts in first place ....

Share this post


Link to post
Share on other sites
... spend some moment on trying find my answers about the performance issues in multiple threads

then you become aware we know about it ...

just go on thread list (not search, not new posts) and there is this little icon saying "Developer posts in thread") click it

and then next to thread # there is icon which takes you to another developer post and so on ...

hard to try it explain to the very same people who ignored the posts in first place ....

It's because you guys haven't put it on the known issues page of the Arma 3 website. For that to not be on there, it suggests that 1) either the issue isn't important to you or 2) you aren't aware of it. This of course may not actually be the case, but that's the impression it gives.

So I'd say you should add the performance issue to that known issues page, since it's an ongoing issue. If you do, you should be more specific about what BIS devs see as the problem, whether that's a CPU- or GPU-utilization problem, or whatever else you guys might think the issue is. People want to see that you are working on it, and they want to see you acknowledge that it's an issue, publicly.

Edited by antoineflemming

Share this post


Link to post
Share on other sites

The funny thing is that Dwarden calls someone else a troll all while trolling away. "Hey guy's, I answered the problem somewhere, just do a search through every developer post since the beginning of time and I'm sure you will find it, oh btw TROLOLOLOLOLOL".

You've known about the issue since what like 2006? Question is, what's being done about it? When you have answers like the one above that sound like you're just trolling along, doesn't really explain much. Just makes you look like you don't know how to fix it and you don't care enough about the problem to fix it. When you see a developer say "Hey we're gonna stick with 32 bit and multi-threading and parallel processing is for chumps, well you kind of get the feeling that either they don't know how to implement it properly into their engine, or it would require too much work, even after 3 generations of said engine, to actually implement it and you would rather just keep money flowing in than have a properly functioning piece of software.

I've seen nothing but Informative posts by White and while they might not always have the best tone or the most respect for the developers, they are never outright hostile or trolling. He's brought the issue to your attention countless times and tried many different ways to show how this problem affects user's and you simply call him a troll.

One might think that you have no idea what you are talking about and have no other response than to call him a troll. Either way, your responses don't reflect very well on the company you represent.

*edit* btw I still enjoy the post by Dwarden where he compares full multi-core usage by the Frostbite engine as being akin to Prime95 or some endless mathematical calculations in defense of why the RV engine can't fully make use of multi-core processor's.

Edited by Insanatrix

Share this post


Link to post
Share on other sites

some people can't be helped as I see, no matter how times I repeat something it's ignored ...

Share this post


Link to post
Share on other sites
some people can't be helped as I see, no matter how times I repeat something it's ignored ...

then put the info in your sig so it is seen everytime you post.

Share this post


Link to post
Share on other sites

i'm afraid my signature would be over the limit and I would get rightfully banned from forum for oversize spammy signature ;)

just example ...

http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2322387&viewfull=1#post2322387

http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2343613#post2343613

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this  

×