Doubling Down: 2x Gigabyte GTX 960s in SLI Put to the Test

I have tested GPU's such as the GTX 770 with 2GB and 4GB of memory, zero difference at 2560x1600. Still if I happen upon a 4GB GTX 960 I will confirm for you that it is a waste of money.

Yea I also remember the 1 and 2GB HD 6950's reviews (I had two of the 2GB version in Crossfire) as well as entry level cards with no business having more than 1 or 2GB, but even though the 128-bit memory bus is justified due to the compression introduced with Maxwell, I was still curious why the performance wasn't much better than the 760. Call it thinking out loud if you will.

I've been away from nVIDIA cards for a couple years due to money issues and was buying AMD cards because they were cheaper, but after coming back and following Maxwell, I know why the performance may be considered "low" compared to a GTX 760, Radeon 280 or 285X.

Thank you for your reply Steve.
 
HMMMMMMM my 290x CF at 1200 core 1.3v in furmark dont even get 85C+ So how are the cards in your benchmark getting 95 from playing a game?
Also where did the SLI temps go? I seen the 970 sli taking almost same watts as 290s but there not on temp chart?
 
HMMMMMMM my 290x CF at 1200 core 1.3v in furmark dont even get 85C+ So how are the cards in your benchmark getting 95 from playing a game?
Also where did the SLI temps go? I seen the 970 sli taking almost same watts as 290s but there not on temp chart?

Going out on a limb here but I am guessing your R9 290X graphics cards don't use the AMD reference cooler. Also the R9 290X cards won't max out in FurMark, they work harder when gaming.

SLI temps? They are the same as single card temps since we tested on an open test bed. It isn't like we incldued Crossfire temperatures and not SLI :S
 
Going out on a limb here but I am guessing your R9 290X graphics cards don't use the AMD reference cooler. Also the R9 290X cards won't max out in FurMark, they work harder when gaming.

SLI temps? They are the same as single card temps since we tested on an open test bed. It isn't like we incldued Crossfire temperatures and not SLI :S
Ok thanks for the reply I can see why then. G1 nvidia cards 12+% oC, 8))) Wanted to upgrade but the dif is not big enough to fork out the cash.
Lolll amd reference cards vs g1 cards haha. 8)
Was just going by your test set up page where it says all cards are Gigabyte's.
 
Last edited:
They are Gigabyte cards, they are early versions that used the reference cooler.
 
Bit "old" now...
But where would a 960 4gb SLI sit in all this testing??
Asus 960 2gb is $300aud where as the 4gb is only $40 extra.

As I didn't quite have the money to go a 970 at the time, I ended up going with a 960 (and then a 4gb version came on special for at the time was less than $10 more than a 2gb 960)
I had the intentions of going SLI at a later date.
 
I'm getting a GTX 960 along with an i5 6600 and may be planning to get another GTX 960 for SLI. So my question is, would the CPU bottleneck with a GTX 960 SLI configuration?
 
I'm getting a GTX 960 along with an i5 6600 and may be planning to get another GTX 960 for SLI. So my question is, would the CPU bottleneck with a GTX 960 SLI configuration?

1) It wont be CPU limited
2) If you're getting the K one, overclock it to the best of its extent
3) Don't make the double mistake of getting 960s ... SLI even. The 380 is faster in single GPU and dramatically faster in Xfire vs the 960s.
 
Thanks fo
1) It wont be CPU limited
2) If you're getting the K one, overclock it to the best of its extent
3) Don't make the double mistake of getting 960s ... SLI even. The 380 is faster in single GPU and dramatically faster in Xfire vs the 960s.
Thanks for the quick reply. While I agree that 380 performs better, I just don't have the money to buy a higher rating psu and ups should I want to Xfire in the future, hence the decision to buy a 960. Getting a 960 now and save up for a 970 or something more powerful for the near future doesn't seem sensible either, hence the desicion to SLI 960 in the near future.
 
Back