|
Advertisement
|
|
|
| 16-02-2012, 16:02 | #17 | |
|
Registered User
![]() |
Quote:
Still good card for single gpu setup, not south for sli. Plus 6970 has an advantage of eyefinity. |
|
|
|
| 16-02-2012, 18:07 | #18 | |
|
Registered User
![]() |
Quote:
http://www.techpowerup.com/reviews/A..._Cu_II/21.html All the results are within a few watts of each other, with the 570 having a lower maximum. It also uses less power in HD video tests. Here's some other sites which demonstrate similar results. A little higher in the second one but the xbitlabs one demonstrates HD video again. http://www.xbitlabs.com/articles/gra...c_5.html#sect0 http://www.hardwarecanucks.com/forum...review-19.html |
|
|
|
| 16-02-2012, 18:15 | #20 |
|
Now in high definition
![]() |
I've always used Nvidia cards myself and was never happy with the image quality on ATI/AMD cards, any time I had one I didn't like it.
Nonetheless, Nvidia cards have taken a serious slide in reliability. There's the well known GeForce 8 failures, but even recently I bought a GT440 for a PC and started getting "Driver stopped responding and recovered" errors. Replaced the card and it's now fine. My cousin also got a GTX460 for a new build - same issue. Even Firefox's GPU acceleration was causing timeout errors. |
|
|
|
Advertisement
|
|
|
| 16-02-2012, 19:03 | #22 |
|
Registered User
![]() |
It could happen that certain tasks will be better optimised in SLI or crossfire and have lower power consumption. Average power draws could change too and for one or the other and be related to drivers.
The maximum is still going to stay the same (x2) and this is what you should always consider when picking a PSU to suit. It's hard to find a review directly comparing SLI 570 directly to crossfire 6970. It would be foolish to compare two reviews too as they could be measuring them at different points, with different applications and with different motherboards/CPUs/configurations. I find that techpowerup are one of the most reliable for actual graphics card consumption as they test the consumption of the card itself and not the whole system at the wall (which adds in PSU inefficiencies and everything). |
|
|
| Thanks from: |
| 16-02-2012, 19:39 | #23 | |
|
Moderator
![]() |
Quote:
http://images.anandtech.com/graphs/graph4239/36072.png http://cdn5.tweaktown.com/content/3/7/3741_42.png I wonder why there's such a large variance between reviews. |
|
|
|
| (2) thanks from: |
| 16-02-2012, 19:49 | #24 | |
|
Registered User
![]() |
Quote:
|
|
|
|
|
Advertisement
|
|
|
| 16-02-2012, 20:33 | #25 |
|
Registered User
![]() |
Wow, that's a fairly big chuck of a gap.
I'm not sure I'd trust anandtech entirely with some of their charts. I previously tracked some of their figures back to reviews where they had taken them from and it seems that they don't use the same systems to test, and they measure at the wall. I suppose it's a lot to ask them to test them all on the exact same system as they probably don't have the cards in their hands all at the same time. The test equipment that techpowerup refers to is most likely expensive as well. I wouldn't rule out bias in any of these sites either. We'll have to get our own testing equipment to be sure! |
|
|
| 16-02-2012, 20:52 | #26 |
|
Moderator
![]() |
|
|
|
| (2) thanks from: |
| 16-02-2012, 21:26 | #28 |
|
Registered User
![]() |
|
|
|
| 16-02-2012, 21:43 | #29 |
|
Registered User
![]() |
NTMK has 2 6970s, and headshot has 2 gtx570s. I'd donate a PC as a test bench, and throw in my 6990 to show what real power consumption is all about
. this could actually work. sure I've been looking into getting a power meter for a while now. |
|
|