facebook rss twitter

Review: NVIDIA GeForce GTX 680 2GB graphics card

by Tarinder Sandhu on 22 March 2012, 13:00 4.5

Tags: NVIDIA (NASDAQ:NVDA)

Quick Link: HEXUS.net/qabdu5

Add to My Vault: x

Overclocking the GeForce GTX 680

The tricky part in overclocking the GTX 680 is knowing how to marry the power target and GPU and memory offset functions such that the card is working at maximum frequency at all times.

Increase the frequencies without taking the power target into account and the GTX 680 works fine, though the TDP limit may stop it from jumping up to the best speeds. So, overclocking is a two-fold process; you increase the frequencies to what you believe the card is capable of and then raise the TDP (power target) accordingly. We believe this is exactly what partners will do when launching pre-overclocked models, that is, have a revised vBIOS with higher settings.

Here, the card is provided with an extra 15 per cent power, the GPU clock is offset (increased) by 150MHz, while the memory is upped by an effective 600MHz, to 6,608MHz.

We can't categorically say what speed the GPU is running at; it continues to jump around, and increasing the memory speed factors into just how much the GPU increases by - the total power is taken into account. We could write more about the mechanism, but the bottom line is that the NVIDIA-provided GTX 680 overclocks up to 1,250MHz core and 6,608MHz memory.

The overclocking didn't show much of an increase in our usual Aliens vs. Predator benchmark hence the Crysis 2 results. The scaling amounts to almost 15 per cent above a stock card. We wait to see how brave partners are with their frequencies.