Measuring temperatures is all about being consitent, therefore we keep a steady eye on monitoring the ambient room temperature to make sure that it stays the same. While this is constantly being monitored, we measure the idle temperature of the card using HWMonitor over a 15 minute period. Once this has been recorded, we set FurMark into motion for 15 minutes and record the results again.
Another key area focussing on the cooling performance is of course temperatures and here we see the card finishing up in the middle of our charts at 68 degrees load. The interesting thing for us however was seeing the very low idle result of 29 degrees which is the lowest result we’ve ever seen in our graphics card tests.
It’s really impressive. I thought the narrower memory bus would strangle it but it’s not to be. Makes you wonder why people would spend the extra scratch on a GTX 670
because when you get a GTX 670 you can also overclock that. Most reviews tend to go for the “if you overclock this card (the GTX 660Ti in this case) you can catch up with the next one in the series” (the GTX 670) but then if you had the next one in the series (the GTX 670) you could overclock that too and reach the next one in the series (GTX 680)..and the cycle continues…
I know. It’s a vicious cycle. I was sceptical of my GTX 660Ti because of the narrow bus but I decided on it just via reading reviews. I’m glad I did & saved the extra money I would’ve spent on a GTX 670.
Ryan has a point with the overclock thing. I was waiting for the EVGA GTX 670 FTW but since then I got a EVGA GTX 660 Ti (standard edition) and I’m glad I saved my self half the cost, there isn’t a single game I can’t max out on everything. I only run 1920×1080, but check another one of these into SLI and you’ve got a system to be proud of.
For goodness’s sake, please change the COLOR of the graph, and highlight the scores for the card you are reviewing against its closest competitors. It was PAINFUL to read those graphs. Seriously.
Otherwise a decent review.