True, but the difference is smaller than it's ever been. 690 is very very close to 2x680.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
SLI/Crossfire @ 2560*1440 and 3240*1920 (XBit Labs)
- Thread starter Final8y
- Start date
True, but the difference is smaller than it's ever been. 690 is very very close to 2x680.
Oh the days before the GHz editions were so boring, now were back to max oc v's max oc.
+1
Most just look at the fps and jump on the bandwagon.
http://techreport.com/articles.x/22890
Not many(if any) can tell it like techreport.
+1
Most just look at the fps and jump on the bandwagon.
http://techreport.com/articles.x/22890
Not many(if any) can tell it like techreport.
Last edited:
+1
Wow missed that one - that is actually a very nice review on the face of it, not had time to read it in depth.
Yes good review methology from techreport, glad you got to the proper review since I linked the discussion instead
, proper link put up now.
I noticed the frame time graphs back when the 560 448 was released, just goes to show how much fps results can be misleading.
You can have the highest fps going, but it's of no use whatsoever if it isn't output in a smooth fashion.
As you can see in the review, both Nvidia and Amd driver teams can push fps figures up whether it be to compete or to be the performance leader in an individual title at the cost of quality.
Pity there isn't more sites reviewing gpu frame times as the vast majority are all fascinated with fps figures even though it doesn't show the whole true story regarding performance.
, proper link put up now.I noticed the frame time graphs back when the 560 448 was released, just goes to show how much fps results can be misleading.
You can have the highest fps going, but it's of no use whatsoever if it isn't output in a smooth fashion.
As you can see in the review, both Nvidia and Amd driver teams can push fps figures up whether it be to compete or to be the performance leader in an individual title at the cost of quality.
Pity there isn't more sites reviewing gpu frame times as the vast majority are all fascinated with fps figures even though it doesn't show the whole true story regarding performance.
Those frametime graphs are quite interesting, because although the nvidia solution has generally less microstutter the 7970 CF setup has enough of a performance advantage that you could artifcially lock the framerate to cancel most of it out without having to sustain too low an fps while the other setups wouldn't quite be able to maintain that framerate limit 100% of the time. With some tweaking atleast for BF3 you could probably get the most regular frame output by a slight margin from the 7970 CF setup while still holding a high enough framerate to not be bothered by input latency:
To quote from the review:
"Notice the disparity between the FPS average and the 99th percentile frame times. Although the 7970 CrossFire config is far and away the at the top of the FPS charts, it's only slightly quicker than the GTX 690 at the 99th percentile. Why is that?"
Last edited:
They don't mention it because anecdotal evidence doesn't mean anything. AMD's apparent awful drivers are never mentioned in reviews because it's simply untrue. What you had was a bad experience with drivers. Full stop. People have driver problems regardless of what brand hardware they're using, and reputable websites never continue this myth of AMD drivers that never ever work.
Law of averages on drivers. Of course they're not all bad, I hardly think that's the point. Take a punt based on what you feel is going to serve you better. AMD cards have a reputation (from end users and people who have switched) of having less than decent crossfire support early on before firming up later on. Tri-fire is less robust.
It's not about looking at one driver issue and offsetting it against one from the other side.
It's not about looking at one driver issue and offsetting it against one from the other side.
Trifire has awful support. ATI drivers have been broken since 12.2 in this respect.
Reviews site do mention if they are having issues with drivers and the games they are testing them with, they didn't mention because they didn't have issues with the games they tested on the rig they were using.
Indeed AMDs Multi GPU does take a while to get off the ground and this time compounded with learning GCN, its been a while and it looks like it has started to get of the ground as im seeing more and more users being able to run 3 and 4 way in Eyeinity on the 7 series from 12.6-12.7 beta.
Yes good review methology from techreport, glad you got to the proper review since I linked the discussion instead
I noticed the frame time graphs back when the 560 448 was released, just goes to show how much fps results can be misleading.
You can have the highest fps going, but it's of no use whatsoever if it isn't output in a smooth fashion.
As you can see in the review, both Nvidia and Amd driver teams can push fps figures up whether it be to compete or to be the performance leader in an individual title at the cost of quality.
Pity there isn't more sites reviewing gpu frame times as the vast majority are all fascinated with fps figures even though it doesn't show the whole true story regarding performance.
Well that's no surprise when such things have been that way from the beginning when it comes to FPS.
Review sites never tell the full story and neither can they with the non fixed variables of the PC, if it was reviewing games performances on consoles then that's a far easier task to give a more true story.
SLI has been fairly decent for myself but that is only 2 GPU's and the drivers are fairly mature now. I don't know what it was like early on and I don't know what 3 would be like driver wise. As Locky says trifire support is still a little lackin. I've read similar stuff elsewhere.
Which is what i said.
Its clear that its not working as it should for everyone.
But this comment does not paint that picture as i would take it as worse or not working period when in fact CF it better now on 12.6-12.7 beta than in the 12.2 on the 7 series overall regardless if it still broken on his setup.
Last edited:
What's this I've been reading about refresh rate being capped at 100hz in sli mode? Is it true?
Nope.
Also bearing in mind (regarding comment on page 5 about review sites) issues with drivers might not show themselves on a simple benchmark so I don't think this is a barometer of driver robustness at all.
This isn't aimed at anyone in particular just something to bear in mind.
This isn't aimed at anyone in particular just something to bear in mind.
Indeed its not and its not their intention to be so, its a resolution and performance bench.
If it was and driver/stability/compatibly/issue finding bench then the approach would have to be different.
But if they do happen to come across an issue it is usually mentioned.
Last edited: