Benchmark help View previous topic :: View next topic  


NFOHump.com Forum Index - Hardware Zone
Page 1 of 1
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sat, 29th Dec 2012 20:00    Post subject: Benchmark help
Ok so i am contemplating an upgrade to an i5 2500k setup from my Phenom X4 940BE

I've read all the benchmark results about the i5, no denying that it destroys my athlon there. But real world game testing is where it's at, and that's where i need your help.

If anyone can share results from a i5 2500k setup, pref with a 7970 that would be really helpful. I'll list my setup below for reference and if you could post your setup too.

Unfortunately my monitor is only 1680x1050.

Windows 8 64bit
6GB DDR2
Gigabyte GA-M720
AMD Phenom II X4 940 Black Edition
XFX 7970 DD
Dell 2209WA monitor

I'm not sure what game would be best for this test. Maybe an inbuilt benchmark option to keep it consistent. Any suggestions?

Any help much appreciated before i outlay $£
Back to top
DXWarlock
VIP Member



Posts: 11422
Location: Florida, USA
PostPosted: Sat, 29th Dec 2012 20:40    Post subject:
I got a i5 2500k(running at 4.4), with a 7870 that's overclocked, and 8g Ram.

What benchmark you want done?
I got this one, the unity engine benchmark thing:

http://alttabgamers.com/unigine_20121018_1824.html

I should probably rerun it..Ive done some tweaking since then

-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf

Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sat, 29th Dec 2012 20:50    Post subject:
Do you have Batman Arkham Asylum? this has an inbuilt benchmark, i'm just downloading it now.

Could you re run Unigine Heaven benchmark @ 1680x1050 resolution it will match my system.

MSI X570 Tomahawk |Corsair Vengeance LPX 32gb 3600mhz | Ryzen 5800X3D | EKWB Watercooling | Seasonic Focus GX 850 Gold PSU | 4090 Founders | Predator X34P UW curved monitor | Window Pro 10 x64
Back to top
DXWarlock
VIP Member



Posts: 11422
Location: Florida, USA
PostPosted: Sat, 29th Dec 2012 20:52    Post subject:
I don't have batman Sad
And not sure if Unigine would be best come to think of it, since you are keeping the same videocard.
It seems to use all my GPU and like 20% of my cpu.

-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf

Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Back to top
DXWarlock
VIP Member



Posts: 11422
Location: Florida, USA
PostPosted: Sat, 29th Dec 2012 21:04    Post subject:
(edit: linked to the same one as before forgot to copy the NEW link..had to change it Razz )

Here ya go
http://alttabgamers.com/unigine_20121229_1501.html

using these settings:

-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf

Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.

Last edited by DXWarlock on Sat, 29th Dec 2012 21:09; edited 3 times in total
Back to top
MinderMast




Posts: 6172
PostPosted: Sat, 29th Dec 2012 21:04    Post subject:
Yeah benchmarks like that are less CPU bound I think.

I still have my BF3 MP testing up here:
http://www.nfohump.com/forum/viewtopic.php?p=2015613&highlight=#2015613
That is with a 2500K @ 4.3

I can test Batman, as well if you want, since I have it installed at the moment.
Edit: Although Asylum is not that demanding, to be honest. Arkham City might be a better fit. What settings you want me to run it at?
Back to top
DXWarlock
VIP Member



Posts: 11422
Location: Florida, USA
PostPosted: Sat, 29th Dec 2012 21:07    Post subject:
I know GTA4 got a HUGE boost is FPS when I got the 2500k. I was nearly unplayable on my old Intel q9450.

And games like forged alliance, and Civ 5. are a HUGE difference too when it gets busy.

-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf

Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sat, 29th Dec 2012 21:10    Post subject:
Here's my Heaven benchmark.

 Spoiler:
 


MinderMast that would be great if you could test Batman for me. Don't rush mate i'm getting my boy to bed now and it will be tomorrow before i can test.

@DXWarlock Thanks

EDIT: just notice tessellation was disabled and aa, i will retest.


Last edited by Grale on Sat, 29th Dec 2012 21:12; edited 2 times in total
Back to top
Frant
King's Bounty



Posts: 24650
Location: Your Mom
PostPosted: Sat, 29th Dec 2012 21:10    Post subject:
Check Anandtech's review for the 2500K. It doesn't really matter which GPU they'll use since it's relative compared to pure 1-2 threaded CPU performance.

Some games are GPU-bound and show smaller differences, some games are CPU-bound and show bigger differences but most often the CPU isn't particularly important in games unless it's a CPU heavy game and most games today are GPU-bound.

Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn!

"The sky was the color of a TV tuned to a dead station" - Neuromancer
Back to top
MinderMast




Posts: 6172
PostPosted: Sat, 29th Dec 2012 21:22    Post subject:
Frant wrote:
Some games are GPU-bound and show smaller differences, some games are CPU-bound and show bigger differences but most often the CPU isn't particularly important in games unless it's a CPU heavy game and most games today are GPU-bound.

Well if you look at some recent game releases, you might see a whole lot of people complaining about sluggish performance, especially open-world games, where it was the CPU that they all wanted. Although this might have more to do with the number of cores available, and not the the CPU itself.

Either way, I think the biggest benefits might not be so easily seen on the benchmarks. A better CPU can add to a better overall "smoothness" in performance. It doesn't always translate in a higher framerate.

Tested Arkham Asylum out of curiosity for now... 250+ FPS on average with maximum somewhere over 360 or something Very Happy

@Grale: I will need exact settings and GPU clocks, so we could make valid comparisons, if you are still up for it later.
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sat, 29th Dec 2012 21:31    Post subject:
Ok Heaven benchmark with DXWarlock settings:



My CPU overclocked @ 3.7ghz (stable)
My GPU overclocked @ Coreclock 1050 Memoryclock 1425

@MinderMast: I will have to download Asylum as i was downloading City by mistake! Will repond tomorrow thankyou Smile


EDIT: All my scores are higher but minimum FPS is terrible! CPU maybe
Back to top
MinderMast




Posts: 6172
PostPosted: Sat, 29th Dec 2012 21:34    Post subject:
Well, City might be better, since its performance might be more reasonable (as in not through the roof Smile). I have that installed as well. It should have more impact on the CPU too.

Edit: While the CPU does have a bigger impact on minimum FPS, from what I understand, the Heaven benchmark has this moment where at the very beginning the FPS is quite low for some reason. You don't see it - it's like the very first frame. For me it's 30-40, but it never goes below 60 after that, so I would suggest you look at the framerate while the benchmark is running to get a better picture.


Last edited by MinderMast on Sat, 29th Dec 2012 21:37; edited 1 time in total
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sat, 29th Dec 2012 21:35    Post subject:
MinderMast wrote:
Well, City might be better, since its performance might be more reasonable (as in not through the roof Smile). I have that installed as well. It should have more impact on the CPU too.


Ok we will stick to City then Very Happy i have it 50% downloaded
Back to top
MinderMast




Posts: 6172
PostPosted: Sat, 29th Dec 2012 22:13    Post subject:
Okay, did a quick run of the becnhmark with Fraps in the background to get a better idea of the performance.



This is with your GPU clocks, on a 2500K @ 4.3 and with these settings for "high":

"low" is the same except AA and Tesselation are completely disabled.

You would normally look for CPU limits by using lower settings, but it was already running into the hundreds, so I left it as it is.

Also, if it wasn't obvious, those are framerate numbers on the Y axis, recorded from start to end on the built in benchmark Smile
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sun, 30th Dec 2012 10:18    Post subject:
Okay, done a test but how did you display that graph from fraps readings?

I'm getting stuttering during the benchmark and that is causing the fps to drop to either 0 or at least half of average fps.

I might be mistaken but it definitely seems to happen when changing scenes or possibly loading.

EDIT: Batman

Min fps: 00
Average fps: 53
Max fps: 60
Back to top
MinderMast




Posts: 6172
PostPosted: Sun, 30th Dec 2012 11:04    Post subject:
Batman does get some stutters for me as well when it changes scenes, nothing too bad, but it can be noticeable. This is why I created a graph to have a better picture of how the game performed.

You need to set the setting before you run the benchmark (compare with my settings) - your max FPS is 60, so I am thinking Vsync is on, for example Smile

As for the graph... Fraps has a benchmark feature (in the FPS tab). When you run it, it will log the data and you can check what data to log. You can choose "FPS", which will log the fps values (each second I think) and then I simply used Excel to build a graph from it.
There might be a more "pro" way to do it, but this was the simplest for me Smile

It's also not entirely accurate, since the framerate can fluctuate wildly even within that one second, but I think it's good enough to get a general idea of the performance.
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sun, 30th Dec 2012 11:44    Post subject:
Yeah how did vsync get enabled!

Ok:

Average: 82
Max: 60
Min: 18

I get stutters throughout the benchmark. First noticable at very start when it pans up from the first guys. Must be something wrong!

I'll try and make a graph somehow i haven't got Excel

what are your results from Batman?
Back to top
MinderMast




Posts: 6172
PostPosted: Sun, 30th Dec 2012 11:57    Post subject:
Min: 29
Max: 150
Avg: 106
On what I called "high".

As you can see from the graph, that minimum framerate is due to that giant dip in the second half of the benchmark. It happened quickly enough that it didn't actually register so low in the frame logging (since it was over within a second) and got clipped at 40. It never even went below 60 if not for that moment.

Stuttering can be due to streaming lots of data. CPU and your HDD probably have an effect on this. You could try lowering some settings (like detail level) until it runs smooth, if you can get to that point, and then we can test again with your settings.

Edit:
Oh, and about the graph... you can run the benchmark with FRAPS (I started it by pressing F11 the moment I saw the in-game benchmark actually running and ended it a second or two after the result appeared in-game), open the created file with notepad and post the values in a spoiler here. There should be around 80 values or so.
Then I can just put it in mine and post a screenshot.

Although it's obvious simply from the numbers already that your performance is lower for whatever reason Smile
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sun, 30th Dec 2012 13:50    Post subject:
When you say high settings do you mean these below, as i used the same settings?

 Spoiler:
 


Ok here's my results: http://www.filedropper.com/benchmarks

I wonder if my setup is really holding back the gpu so much.
Back to top
MinderMast




Posts: 6172
PostPosted: Sun, 30th Dec 2012 14:02    Post subject:
I am afraid there is something wrong with the benchmark files - it seems it only ran for 12 seconds Smile

And yes, those settings in the screenshot are what I used as "high".

One other thing: is your game updated? There were some serious issues with DX11 performance in Arkham City when it was released. They fixed it later with some patch.
If that is the case, you can try disabling DX11 features in the settings.
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sun, 30th Dec 2012 14:42    Post subject:
Ok Game updated, which wasn't and ran the benchmark test properly this time.

New benchmark: http://www.filedropper.com/benchmarks_1

MIN:7
MAX:147
AVG:86

Thanks for your help MinderMast much appreciated

MSI X570 Tomahawk |Corsair Vengeance LPX 32gb 3600mhz | Ryzen 5800X3D | EKWB Watercooling | Seasonic Focus GX 850 Gold PSU | 4090 Founders | Predator X34P UW curved monitor | Window Pro 10 x64
Back to top
MinderMast




Posts: 6172
PostPosted: Sun, 30th Dec 2012 15:16    Post subject:
Well, here is a nice comparison graph for you:
 Spoiler:
 

Okay here is the "improved" version Smile :
 Spoiler:
 

I ran my benchmark twice to be sure - results were pretty much identical, so it should be good enough for comparison.
Just by the numbers it seems like there is a difference of 15-30 FPS depending on the scene. It's still a question of how noticeable this is in-game though.
No doubt that the CPU is holding your GPU back, but whether it's something that needs to be dealt with right away, is something for you decide Smile
Back to top
Mr.Tinkles




Posts: 12378
Location: Reino de Suecia
PostPosted: Sun, 30th Dec 2012 15:19    Post subject:
I like the first graph better... Very Happy
Back to top
Grale
Banned



Posts: 3321
Location: Invert
PostPosted: Sun, 30th Dec 2012 15:27    Post subject:
Prefetian wrote:
I like the first graph better... Very Happy


Very Happy

I'm going to put my card and cpu back to stock and run some more comparisons. I'll download Excel so i can run that graph as it's very helpful.

I'm also going to re install my windows 7 true image and see if it's anything to do with w8.

I'll post my findings MinderMast see what you think, but i'm starting to think a cpu, mobo and memory upgrade might be on the cards.
Back to top
MinderMast




Posts: 6172
PostPosted: Sun, 30th Dec 2012 15:42    Post subject:
You should also really check out out that link Frant posted earlier, if you haven't already.
In some games the difference can be huge.
Back to top
Stige




Posts: 3545
Location: Finland
PostPosted: Mon, 31st Dec 2012 11:28    Post subject:
You can find a few game benchmarks in the comparison here aswell: www.anandtech.com/bench/Product/80?vs=288&i=50.49.48.47

I could do some benchmarking for you but I'm running Crossfire atm and I had to squeeze the cards in the case so too lazy to take them off lol (Sonata III, super tight fit)
Back to top
Page 1 of 1 All times are GMT + 1 Hour
NFOHump.com Forum Index - Hardware Zone
Signature/Avatar nuking: none (can be changed in your profile)  


Display posts from previous:   

Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB 2.0.8 © 2001, 2002 phpBB Group