So I've recently come by a reasonably large sum of money and decided that some of it should go into upgrading my already considerably fast computer.
My rig is currently:
8Gb 1600 RAM (2x4Gb)
Asus P8P67B3 Motherboard
Intel i5 2500k (OC'd to 4 GHz)
XFX HD 7870 Black Edition
600W Corsair PSU
I have two concerns: a) Will my PSU handle the second GPU without problems, and b) Would the second card be a valuable investment for later on, because the next time I will be able to upgrade my system is somewhere around this time next year.
Thanks in advance!
EDIT: I forgot to add in what I would be using these cards for, I would be generally be just using them for playing video games, especially heavily modded games like Skyrim.
Thanks for the input, but do you think that the upgrade would be worth it, as I aim to hit ultra/very high on the most demanding games that would be releasing next year. It's doable at the moment, but I currently have no knowledge of projected requirements for next year's tech.
I also think that a PSU upgrade would be doable, if absolutely necessary.
There's a single 7870 in there too, not doing too well though. Are you saying you have double the CPU that Tom's was testing with and it's strange that that's why you're getting more performance? Or strange that dual GPUs would make performance so inconsistent? Stutter / performance inconsistency has always been a characteristic of SLI/Crossfire. And people often compare their own results to benchmark results they see online, and they're almost always wrong to do so.
Firstly, is the rest of the spec the same? More importantly, are you using the exact same settings, both in game and in Catalyst control panel? And also very important, are you testing the same part of the game they tested? You know that some parts of games deliver massively higher framerates than others. For an extreme example, comparing the interior of a tiny storeroom with a wide open outdoor area full of enemies and combat.
I was saying that it was strange that the performance was inconsistent, and even dropped below the performance line of the single card. I believe I am using the same settings, my resolution is the same and I used the same preset settings in-game, but I didn't change anything within catalyst. I'm going off of the fact that I went into a large camp the other day guns a blazing without the fps dropping below 40 or 50.
Oh I'll tell ya what actually, I recall somebody saying earlier today that Far Cry 3 had major performance issues before a certain patch. Is your version patched? That might actually be why Tom's got pretty poor performance. If it is, they should probably re-run those tests. According to the review, you'd need a GTX670 to max it at 1080p. As for the inconsistency though, that is something that's been seen on SLI and especially Crossfire in many games before. There are some interesting articles about it (it's called 'microstuttering') - I'll see if I can find one.