Nvidia GeForce GTX 680 – The Nitty Gritty

Last of all, this section is one that I’m sure a number of people will be keen to look at. For those out there who are Nvidia fan boys and have experienced overclocking on their cards in the past and having to raise the shader clocks independently to the core and memory clocks will now see a whole now way of overclocking. Gone are the days of the shader clock and now overclocking only involves the core and memory clocks along with adjusting the cards overall power target.

The GTX 680 as we have seen back in the GPU Boost section, works within a TDP range at stock and so when the boost comes into factor, the total draw of the card is still under the 195W maximum.

With the likes of the new EVGA PrecisionX v3, users can overclock in a number or ways. The first of which is by simply increasing the cards TDP range. By doing this you are allowing the card to dynamically overclock itself as part of the GPU boost but at a higher range. This is the simplest and easiest form of overclocking and a TPU increase of upto 32% is guaranteed to work with the PrecisionX software.

The second option is to give an offset to the cards GPU clock. By doing this you are still working within the stock power target and if the card runs out of power to run at the higher frequency it will clock itself down to stay within the power limit. The great thing about this is that it still works in coalition with the GPU boost technology which will still raise and lower the clocks as required.

One step further than this is to adjust both the GPU clock offset and the power offset to allow the card to run at higher clocks within a greater power envelope.

As part of our meeting with Nvidia, one of their representatives Tom took some time to explain this all as part of a live demo which you can find the video to in our conclusion.