Nvidia GeForce GTX 750Ti Performance Review: Overclocking!

I have done a brief overview of the Palit GeForce GTX 750Ti StormX Dual 2GB graphics card under Ubuntu and Windows:


Now, it is time to do some overclocking! I will be using EVGA PrecisionX to overclock, and PerformanceTest 8.0’s 3D Graphics suite to benchmark.

Something to note: this card is factory overclocked (details here). This means the card runs at higher frequencies than Nvidia’s reference design card. I firstly underclocked this card to match Nvidia’s reference clocks:

PrecisionX REF

The card scored 3578 in Passmark GPU tests. Now, to run the card at its own factory speeds:

PrecisionX Normal

At factory speeds, the card scored 4020 in Passmark! That’s a 12% improvement, which is a fair bit for factory overclocking (my Sapphire HD5750 VaporX was also factory overclocked, by a measly 10MHz).

Time for some overclocking! Let’s try a 100MHz overclock on GPU and VRAM, and set the temp target to 90 degC:

PrecisionX +100

After that, Passmark gave us 4281 points. The highest Boost Clock I was was 1446MHz! I soldiered on to find the highest possible clocks, and I came up with:

GPU: 1347MHz (+145 MHz, +12% over factory)
VRAM: 1565MHz (+125 MHz, +4% over factory)
BOOST: 1491MHz max
Passmark points: 4364 points (+344, +8.5% over factory
Max Temperature: 57 degC.

PrecisionX MAX

So there you have it! I was able to squeeze another 8.5% out of the GPU. The dual fan cooler kept the GPU under 60 degrees, even under overclocking!

I also ran 3DMark with a slightly lower overclock (+75MHz on GPU and VRAM), and got the following results:

750Ti OC

750Ti Graph OC

Compared to the results from yesterday, this is a score increase of 65 points (about a 1% improvement). Under these settings, I was able to get the Windows 7 Experience Index to 7.8 (from 7.7).

I wonder if you can do some overclocking in Linux? Hmmmm…..

Nvidia GeForce GTX 750Ti Performance Review: Windows!
I am terrible at Kerbal Space Program: Episode 1

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.