header photo
Showing posts with label overclocking. Show all posts
Showing posts with label overclocking. Show all posts

Wednesday, January 27, 2010

Intel Core i5 750 Overclock with Gigabyte QuickBoost

I dove into the realms of overclocking today with Gigabyte's included overclocking tool that came with my motherboard. It is a very easy way to overclock although I could see one getting into trouble by using this utility. The tool has three settings: faster, turbo, and twin turbo. Due to the fact that I still only have the stock Intel cooler, I limited myself to the faster and turbo options.

Gigabyte QuickBoost Utility

  • Baseline
    • Base Clock: 133MHz
    • 4 thread max multiplier: 21
    • CPU clock: 2798 MHz
    • CPU voltage: 1.20V
    • 100% load temp: 64
    • Intel Turbo mode: on
    • Memory clock: 1333MHz
  • Faster
    • Base Clock: 140MHz
    • 4 thread max multiplier: 21
    • CPU clock: 2940 MHz
    • CPU voltage: 1.20V
    • 100% load temp: 66
    • Intel Turbo mode: on
    • Memory clock: 1400MHz
  • Turbo
    • Base Clock: 150MHz
    • 4 thread max multiplier: 21
    • CPU clock: 3150 MHz
    • CPU voltage: 1.23V
    • 100% load temp: 69
    • Intel Turbo mode: off
    • Memory clock: 1500MHz
  • Twin Turbo (Did not test)
    • Base Clock: 200MHz
    • 4 thread max multiplier: 19
    • CPU clock: 3800 MHz
    • CPU voltage: NA
    • 100% load temp: NA
    • Intel Turbo mode: off
    • Memory clock: NA

I was nervous about selecting the different settings since the utility would not provide voltage and memory clock information. The turbo mode took my memory from its stock 1333MHz to 1500MHz. Although my memory happens to be able to run at those speeds, I did not know where it would end up until the computer restarted with the settings saved to the bios.


Supreme Commander Benchmark
I ran through the Supreme Commander benchmark at the baseline, faster and turbo settings to see if any gains in performance can be seen. I ran the benchmark at 1920x1080 resolution. Results are below.

Unfortunately, it seems that no gain in performance can be seen according to the average frame rates. It is interesting how the baseline settings resulted in the highest minimum frames per second. I do not understand this at all. It seems though that this Core i5 750 can handle the stress of this game very well.

Sunday, January 24, 2010

MSI ATI 5770 Overclocking, 3DMark Vantage, S.T.A.L.K.E.R.: Call Of Pripyat Benchmark, Supreme Commader, DiRT 2

I explored overclocking a MSI ATI 5770 seaweed edition with the use of MSI's Afterburner program, MSI's Kombuster GPU stresser (based on Furmark)  and 3DmarK Vantage on the performance setting. Unfortunately I currently do not have any game benchmarks to run yet.

The stock GPU clock rate is 850MHz with a memory clock rate of 1200MHz. I explored stability with the Kombuster program and ran 3Dmark Vantage at two different overclocks. The highest overclock that I reached that seems to be stable is a core clock of 950MHz and a memory clock of 1400MHz. Temperatures never exceeded 65C and the fan only reached 41%. The maximum that the afterburner tool can overclock to seems to be a core clock of 960MHz and a memory clock of 1445MHz. While stressing the GPU at this level, I experienced stability issues; screen started flashing and the mouse pointer became garbled.


3Dmark Vantage
3Dmark Vantage on the performance setting scores, as well as the individual CPU and GPU scores at baseline and two overclocks.


The baseline 3Dmark score was 8365 while the highest achieved overclock score reached 9222 resulting in a 10% increase.

S.T.A.L.K.E.R.: Call Of Pripyat Benchmark
First game benchmark used is S.T.A.L.K.E.R.: Call Of Pripyat Benchmark. Resolution is 1920x1080 using DX 11 and on the medium preset. Since the benchmark reports four different cases, the absolute minimums are presented below while the averages represented below are averaged evenly across all four cases resulting in the numbers below.

Unfortunately, according to this benchmark, overclocking might not be worth it. Average frames rates are already very playable and overclocking only adds ~5 FPS. Overclocking only increases the minimum frame rate by 1.5 FPS.

Supreme Commander Benchmark
I recently acquired supreme commander and ran the benchmark at the three overclock levels I have achieved with my ATI 5770. This game is highly CPU dependent for the simulation and AI calculations needed.
 
The results seem to show that this graphics card will run Supreme Commander fine without the need to overclock. This game is definitely limited by my CPU even though it is a core i5 750. It will be interesting to see how the results scale with CPU overclocking or increasing the resolution and image options.

DiRT 2 Benchmark
I finally got around to using DiRT 2 to benchmark the various overclock levels that I have achieved. DiRT 2 came with my graphics card. The game is not appealing to me and since it came with my card, I have to use Steam. I find Steam annoying and will promptly uninstall it after I am finished using DiRT 2 as a benchmark.

 The benchmark was set on a resolution of 1920x1080 with 4xAA and everything else on the "high" setting. The results are shown below.


 The game is very playable with the minimum frame rate never dropping below 30 FPS. Overclocking does improve the average frame rate by 4 FPS but since the game is already above 30 FPS, there is no difference in game play. Overclocking this card seems to result in little benefit.
 

[H]ardOCP's ATI 5770 overclocking results concluded with a maximum  core of 960MHz and a memory clock of 1400MHz. My results came fairly close to this. Tom's Hardware results only hit a core of 895MHz and a memory clock of 1430MHz.