Overclock.net banner

CPU and GPU OC resluts for Unigine Heaven

5K views 4 replies 3 participants last post by  CBZ323 
#1 ·
Hello everyone,

I was curious (read:bored) about what the effect of overclocking GPU and CPU was in a banchmarking application. This is my first time performing a test like this so please bear with me if I have made any mistake or have skipped over some impoprtant aspect.
There are many applications and each one utilizes different resources from CPU and GPU but I decided to use Unigine Heaven as a benchmark.

heaven.jpg

I tested three Intel i5 2500k CPU clocks: 3.33 , 4.0 and 4.5 and just one GPU OC since i am currently in a very hot room and I dont want to push my GPUs to extreme levels going from stock sttings in my two PNY GTX 560 Ti in SLI of 822/1645/2004 to 900/1800/2103

I only ran a test with each setting once, so it is not a very scientifical method but seeing the results it looks like it didnt really matter.

Here are the settings for the test:

hardwarep.png


And here are the results in all six combinations of my CPU and GPU overclocks:

testxhg.png


Even though I knew that Unigine Heaven is a very GPU intensive benchmark I thought a higher CPU might help scratch a few FPS.
teaching.gif
It turns out thar the effect of CPU OC is absolutely zero. There are some minor fluctuations but as I mentioned before, i only ran each test once, so if I did each test 3 or 5 times the results would have probably been almost identical accross the every CPU setting.

On the other hand, OCing the GPU will give us an increase of 6-7 fps in average. which is pretty sweet considering the heat was only slightly superior to stock settings.

As a conclusion, this mini test confirms the theory that as long as your CPU is not bottlenecking the GPUs there will be no substantial changes in performance in very GPU intensive applications such as Unigine Heaven.

Please take into account that this doesnt mean that other applications/games such as Bad Company 2 or Metro 2033 might not benefit from a higher CPU clock.

Please comment and if you have any suggestions on how to improve an amateur test like this or want me to look into something that I havent gone through or analyzed in this test I would me more than glad to do it!
wink.gif


Cheers!
cheers.gif
 
See less See more
6
#2 ·
Something wrong with those resutls. CPU @4.0Ghz performs considerably worse than @3.33...
 
#3 ·
Quote:
Originally Posted by ronnin426850;14244431
Something wrong with those resutls. CPU @4.0Ghz performs considerably worse than @3.33...
As i mentioned I only ran the test once, so if I run that many times it would probably be very similar to the others. The only considerable difference in my opinion is the min FPS, which if I remember correctly it happened when it changed to a different sequence. The average is pretty similar as well as the max but i agree that to my surprise the 4.0 performed lower in every aspect than the 3.33
 
#4 ·
Bah. The differences in results you got in different cpu speeds are within margin of error.
Couple weeks ago I too tested hwbot heaven extreme preset with my cpu at 3.8 and 4.8 speeds. No change in points whatsoever.
=heaven is gpu intensive, as we alrdy know.
Maybe you should do some benchmarking with BFBC2 multiplayer at some 32player map, there the cpu speed differences should REALLY start to showup.

I could do some BFBC2 testing too on my sig rig and put results here.
Currently on a vacation though so..
(I'll be getting better cooler for my gpu so I can oc it to its max, and I'll continue testing my 2500K at higher clocks when I get back.)

Sent from my HTC Desire using Tapatalk with this enabled just to annoy you.
 
#5 ·
Quote:
Originally Posted by lagittaja;14251311
Bah. The differences in results you got in different cpu speeds are within margin of error.
I dont know if you bothered to read the text but that's what i wrote in the review.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top