So in my computer plan it has a 7970... And in the future i want to add a second 7970. I have been hearing alot about problems with crossfire but not SLI. Do these problems happen alot? How do you fix them? I might get a 680 instead....
any answers?
If you hear about someone not being able to get crossfire working, that usually means the motherboard didnt support it in the first place...
No, Nvidia is not safe either. And it is most often a problem with the application you are running not the cards themselves. [Also mobos, like RyTak said] Meaning some games will not even look for the 2nd card and it won't be used.
What is a more common problem is Scaling. When 2 cards are in SLI or Crossfire they hold hands and share the load. 2 cards are better than 1 but will rarely provide you with 100% increase over a single card. The technology is mature enough these days where both Nvidia and AMD scale about 90%+ (this will vary on the game)
To directly indirectly answer your question. 7970s CF should scale very well.
The recent issue people bring up is related to frame latency. That is, the difference in the amount of time it takes to render one frame. On any combination of GPU/s, not every frame will take exactly as long to be rendered, for various reasons. I'm uncomfortable saying it's impossible, but it might as well be.
That said, single cards generally have negligible (unnoticeable) frame latencies, while multi-GPU systems tend to have more. Sometimes it's noticeable, sometimes it isn't. That depends on what's being displayed, what monitor it's being displayed on, and who's watching it. Measurements of frame latencies show that Crossfire generally has more latency than SLI, but there is debate (and mostly a subjective debate at that) over whether the difference is noticeable. Reviews measuring latency usually just pick an arbitrary point where they say it becomes noticeable, and count how many times each setup crosses that point.
Some people swear that Crossfire is totally unusable because of it, and some people don't see a difference. I'm going to groundlessly assume, until I see a study or something showing otherwise, that most people wouldn't notice a difference.