AMD CPU Bottleneck?

[oktyabr 12]

Because you didn't qualify your statement.  Is an Nvidia MX440 better for ARMA3 than an AMD R9 390?  Your argument is invalid.   I've had both, at the same time, many times.  

 

[glocks999 0]

A bit off topic but im new to the forums why cant i post?

[St. Jimmy 272]

A bit off topic but im new to the forums why cant i post?

Because of spam bots you need to post couple times before you can make a topic and anything like that.

[chrisb 196]

 

 

"For ARMA 3, Nvidia performs much better than AMD (I know because I have both)."

 

 

 

To be honest this statement is just.. well..

 

Problem with statements like yours is, and this is the main point... Its your opinion.. Which in most cases, means nothing to other players.

[war_lord 934]

But in game performance is an objective thing, frames are frames, that's not an opinion, it's a fact. 

[domokun 515]

Rather than post subjectives statements and declaring them fact, how about posting objective comparisons:

CPUs

http://www.techspot.com/review/712-arma-3-benchmarks/page5.html

GPUs

http://www.techspot.com/review/712-arma-3-benchmarks/page4.html

This article is really what we need.

Although it is 2.5 years old, I believe it is still relevant because although the engine has been patched dozens of times since, I think that it hasn't significantly changed.

And here's a far more recent article but which is less comprehensive (focused on Skylake vs predecessors).

http://www.hardware.fr/articles/940-15/cpu-jeux-3d-crysis-3-arma-iii.html

In short, I would like 2016 to be the year when people remembered that science is a method of investigation,and NOT a belief system

[Brisse 78]

Rather than post subjectives statements and declaring them fact, how about posting objective comparisons:

CPUs

http://www.techspot.com/review/712-arma-3-benchmarks/page5.html

 

Wow. What a ridiculous apples vs. oranges comparison. They take the FX-8350 and compare to the i7 4770k which costs TWICE AS MUCH. They are not even competing in the same marketing segment. Even more ridiculous is that the extra ~150$ that one has to spend on the i7 4770k (does not even include platform cost, that is for CPU only) only manages to chunk out 1 measly frame per second more on average (FX 48fps and i7 49fps) at stock frequency! Hilarious.

[St. Jimmy 272]

Wow. What a ridiculous apples vs. oranges comparison. They take the FX-8350 and compare to the i7 4770k which costs TWICE AS MUCH. They are not even competing in the same marketing segment. Even more ridiculous is that the extra ~150$ that one has to spend on the i7 4770k (does not even include platform cost, that is for CPU only) only manages to chunk out 1 measly frame per second more on average (FX 48fps and i7 49fps) at stock frequency! Hilarious.

The fps increase with the i7 4770K doesn't look right when they overclocked. The chart where it shows other CPUs shows 53fps @ 3.5GHz but the OC chart shows 49fps @ 3.5GHz. Also there something weird because the increase from 4.0GHz->4.5GHz doesn't follow linear path. Something ain't right with those i7 results. They've screwed that big time when you just compare the chart that has all the CPUs below. FX8350 gets the same fps but suddenly i7 4770K has 4fps difference. The OC results are suddenly worse :D

 

Different benchmarks show different results. In this benchmark the difference is bigger http://www.hardware.fr/articles/940-15/cpu-jeux-3d-crysis-3-arma-iii.html

AMD just can't climb on the top with their FX stuff and they don't have any CPU that could compete with the high end Intel in Arma. If you want to compare Intel and AMD in the same price category then just stop that right at the beginning. AMD is good for lower budget category, and Intel if you've more money. For other type of games AMD FX-8350 is a very good choice for the price.

i7 performance is pretty much the same with i5. The stock clock speed is just usually 0.1-0.2GHz higher in i7 and that's why it gets 0.5-2fps more. Give them same clocks and there shouldn't be more than 1fps difference between i5 and i7.

[war_lord 934]

Hang on, are we talking Arma 3 or general gaming?

For Arma III we're talking a clear superiority in favor of an Intel CPU and Nvidia GPU combo in terms of performance. Arma 3 just doesn't work very well on even relatively high end AMD hardware.

For more general gaming, AMD is a perfectly viable choice for a gamer on a tight budget.  

[calin_banc 19]

Rather than post subjectives statements and declaring them fact, how about posting objective comparisons:

CPUs

http://www.techspot.com/review/712-arma-3-benchmarks/page5.html

 

Some of these reviewers really need to play what they're testing or at leas show some interest/knowledge about what they're doing. Infantry scenario is more GPU limited than CPU limited. Look here - http://gamegpu.ru/action-/-fps-/-tps/arma-iii-test-gpu.html. Don't know about frame times, but that old i3 2100 is ~ equal to a fx8350@4,5GHz.

 

The GPUs are fairly equal AMD vs. nVIDIA.

[oktyabr 12]

The fps increase with the i7 4770K doesn't look right when they overclocked. The chart where it shows other CPUs shows 53fps @ 3.5GHz but the OC chart shows 49fps @ 3.5GHz. Also there something weird because the increase from 4.0GHz->4.5GHz doesn't follow linear path. Something ain't right with those i7 results. They've screwed that big time when you just compare the chart that has all the CPUs below. FX8350 gets the same fps but suddenly i7 4770K has 4fps difference. The OC results are suddenly worse :D

 

Different benchmarks show different results. In this benchmark the difference is bigger http://www.hardware.fr/articles/940-15/cpu-jeux-3d-crysis-3-arma-iii.html

AMD just can't climb on the top with their FX stuff and they don't have any CPU that could compete with the high end Intel in Arma. If you want to compare Intel and AMD in the same price category then just stop that right at the beginning. AMD is good for lower budget category, and Intel if you've more money. For other type of games AMD FX-8350 is a very good choice for the price.

i7 performance is pretty much the same with i5. The stock clock speed is just usually 0.1-0.2GHz higher in i7 and that's why it gets 0.5-2fps more. Give them same clocks and there shouldn't be more than 1fps difference between i5 and i7.

I don't think anyone is arguing that *currently* AMD can't keep up with Intel on ARMA3.  This >isn't< the CPU, this is how the ARMA3 game engine is coded.  As I said a page back my son's new i5-6500@3.2 w/ an R9 280 gets almost twice the frames that my water cooled OC'd AMD FX4100 @ 4.4 Ghz does, and that's with a new R9 390!  For all the other games I play, including Squad on "EPIC" and my beloved Assetto Corsa racing sim, on Eyefinity across five feet of monitors (5760x1200) with all the eye candy maxed out I still get 40-60 fps.   Squad is an interesting case because monitoring CPU usage (Process Lasso Pro) shows it makes MUCH more efficient use of CPU cycles and threads than does ARMA3...  So now, if I don't want my ARMA to look like a slide show I get to spend $700+ on a new skylake to be a home for my R9 390?  I wish I didn't love ARMA3 so much...

Was just hoping someone would have popped up and said "YES!  Arma3 can make great use of 8 cores!  Save a few pennies and upgrade your FX4100 to a FX8350..."   :(

Any benchmarks out there showing a broad range of AMD cpus with Arma3????

[vitamin 18]

Was just hoping someone would have popped up and said "YES!  Arma3 can make great use of 8 cores!  Save a few pennies and upgrade your FX4100 to a FX8350..."   :(

Any benchmarks out there showing a broad range of AMD cpus with Arma3????

Upgrading to 8350 won't make much of a difference in A3 unfortunately. This game is poorly optimized/coded and relies highly on IPC and what counts most is pure single thread performance. Check the videos below to see some tests:

 

https://www.youtube.com/watch?v=K4OSA-vd1h8

 

 

 

And here's some good read:

http://www.overclock.net/t/1508003/arma-3-horrible-performance

[St. Jimmy 272]

Upgrading to 8350 won't make much of a difference in A3 unfortunately. This game is poorly optimized/coded and relies highly on IPC and what counts most is pure single thread performance. Check the videos below to see some tests:

 

videos

 

And here's some good read:

http://www.overclock.net/t/1508003/arma-3-horrible-performance

Wow that second video :o That performance is dam good when compared to the FX-8350. You get the Pentium for 82€ and FX-8350 for 209€ in here. Even FX-4300 costs 90€ so seems like if you can OC, the Pentium is a dam good CPU for the price in Arma 3.

[Freghar 73]

Wow that second video :o That performance is dam good when compared to the FX-8350. You get the Pentium for 82€ and FX-8350 for 209€ in here. Even FX-4300 costs 90€ so seems like if you can OC, the Pentium is a dam good CPU for the price in Arma 3.

Indeed, it's been known for some time that cheap Pentium dualcores are great for singlethreaded performance if you overclock them. They don't have huge amounts of cache, but have the price advantage. (Also, games still cannot use the integrated GPU in addition to a dedicated one, so you don't benefit from i3/i5/i7 GPU.)

On the other hand, the AMD Bulldozer (and Piledriver) has its share of (mainly cache) problems, significantly affecting single-threaded performance, so it cannot really compete with Intel in that area at the moment. It will be interesting to see what Zen brings along in 2016 Q4 - wouldn't be the first time for AMD to give Intel a run for the money, but we'll have to see.

[I give up 152]

With all respect for the videos and opinions, these are bit outdated. At time of the videos ARMA 3 performance was better.

Now is like this.

Everything maxed at 4K VD.

 

https://www.youtube.com/watch?v=5GLKuvP0ar8

 

Btw, 12.5 GB of RAM loaded (pagefile disabled).

 

But I would not worry to much about hardware, because soon we face gameplay, hardware does not count.

 

Nice frames on Paros with Editor.

https://www.youtube.com/watch?v=m_IiiYB7QiY

 

Nasty frames in Paros playing Pilgrimage.

https://www.youtube.com/watch?v=R_jzGljfKwE

[cosmic10r 2331]

when you said 4k i thought you meant 4k resolution and i thought... not that bad... but now i see what you mean lol..

 

query

-what is the overlay program you guys are using... is that msi afterburner?

[St. Jimmy 272]

query

-what is the overlay program you guys are using... is that msi afterburner?

Yep MSI Afterburner.

 

I'll post later in this post my fps results in Combined Arms showcase if you want something to compare to Intel 2600K @4.6GHz. I'll also do a second video with my optimal graphic settings. MSI Afterburner recording takes some fps but I use RAM to buffer it and lower quality so the impact isn't so big, maybe 2-5fps.

 

Same graphic settings as bratwurste. I died pretty quickly first so I needed to load a save. That's why there's that very small cut.

 

Bit more optimal settings that I normally use for infantry:

[domokun 515]

when you said 4k i thought you meant 4k resolution and i thought... not that bad... but now i see what you mean lol..

 

query

-what is the overlay program you guys are using... is that msi afterburner?

what does he mean? because I suspect that 4k to most people here means 2160p, 3840 x 2160 or UHD-1.

https://en.wikipedia.org/wiki/4K_resolution

[cosmic10r 2331]

what does he mean? because I suspect that 4k to most people here means 2160p, 3840 x 2160 or UHD-1.

https://en.wikipedia.org/wiki/4K_resolution

 

He meant 4000 View Distance... ;)

[I give up 152]

Yep MSI Afterburner.

 

I'll post later in this post my fps results in Combined Arms showcase if you want something to compare to Intel 2600K @4.6GHz. I'll also do a second video with my optimal graphic settings. MSI Afterburner recording takes some fps but I use RAM to buffer it and lower quality so the impact isn't so big, maybe 2-5fps.

 

Same graphic settings as bratwurste. I died pretty quickly first so I needed to load a save. That's why there's that very small cut.

 

Bit more optimal settings that I normally use for infantry:

Well mate. already figured what are the mine issues with performance.

Shadows, mainly shadow transition.

In these circumstances with shadows Disabled  I have 75 fps (minimum), with shadows in Low i have 30 fps (minimum). Thats a 45 fps loss.

Only changed the shadows from Disabled to Low, all the other graphics settings remain the same.

Take a look.

https://www.youtube.com/watch?v=kySXA9S33DU

Now, how this nasty performance impact is related with "cpu bottleneck"? In my book is related with graphics.

[St. Jimmy 272]

Well mate. already figured what are the mine issues with performance.

Shadows, mainly shadow transition.

In these circumstances with shadows Disabled  I have 75 fps (minimum), with shadows in Low i have 30 fps (minimum). Thats a 45 fps loss.

Only changed the shadows from Disabled to Low, all the other graphics settings remain the same.

Take a look.

Now, how this nasty performance impact is related with "cpu bottleneck"? In my book is related with graphics.

Always use shadows either high or above or disabled. It's a long time know thing that the low and normal shadows are CPU heavy. Low and Normal use Stencil Shadows and that's the cause for that.

Good that you found out that. It's a nasty thing.

[I give up 152]

Always use shadows either high or above or disabled. It's a long time know thing that the low and normal shadows are CPU heavy. Low and Normal use Stencil Shadows and that's the cause for that.

Good that you found out that. It's a nasty thing.

Are you sure mate? Its Low and Normal?

I am having exactly the same fps with shadows in High, Very High and Ultra. 

Means I am losing 45 fps whatever the quality preset, when comparing to having shadows disabled.

Take a look.

https://www.youtube.com/watch?v=a8q0SEBoB14&feature=youtu.be

Are you sure its Low and Normal?

[St. Jimmy 272]

Are you sure mate? Its Low and Normal?

I am having exactly the same fps with shadows in High, Very High and Ultra. 

Means I am losing 45 fps whatever the quality preset, when comparing to having shadows disabled.

Take a look.

the video

Are you sure its Low and Normal?

OK that's interesting. I tried myself and I didn't see behavior you get. That's a huge difference. I think you need to report that to the devs and give some specs. I've only ~10fps difference between shadows on/off in bit same kind of a scenario. I couldn't get exactly the same scenario as in your video but I can now see that there must be something wrong in there. You've 60fps (90->30) drop in there where I got like 20fps (120->100). Could you share the scenario? I could test more precisely what's the difference between us. That fps drop really seems to be huge.

[cosmic10r 2331]

well weird one for ya...

 

I replicated the hunter driveby on altis airport. and i got about the same fps drop when the hunters drove by on every shadow setting including shadows disabled    :confused:

about 10-12 fps.

[I give up 152]

Well, shadows are the fps killer, in my end.

I've tried with everthing maxed (Ultra/8xAA) and with shadows disabled and basically there are no performance issues.

With shadows in Low I have the same perfromance drop that I have with settings in High/4xAA 

https://www.youtube.com/watch?v=aLn4htOnWGU

Starting to think it's related with AMD gpu's only and is also related with the known issues with LOD transition that these gpu's have under DX11.