and ati drivers are complete crap, what r u smoking?
Infracted. Post like these will only lead to AMD vs nVidia brand flamewars. Please refer to the Computer-section rules, point 2.
Last edited by mmoc7c6c75675f; 2013-01-11 at 01:06 AM.
Milk was a bad choice.
2013 MMO-Champion User of the Year (2nd runner up)
No, only that when you added up all of the numbers into one pool, the numbers favor nVidia.
Bear in mind that that sort of outcome doesn't tell us much.
Firstly, there are obvious issues with how the number has been formed. As an example, Call of Duty 4: Modern Warfare has a HUGE bias towards nVidia. Or Civilization 5. Benchmarks like these almost completely invalidate the outcome of "relative performance," because it shows no signs of what you will actually see in day to day scenarios unless you play exactly the same titles in equal amounts.
Secondly, this was 6 months to 8 month ago. Especially as of late, AMD have been pushing pretty huge improvements across the board, as have nVidia. For anything resembling reality, new benchmarks are required.
---------- Post added 2013-01-11 at 01:26 AM ----------
Don't pull that sort of bullshit. I gave you those numbers. You did not make any argument with any kind of information, and claiming anything else is frankly offensive.
http://www.techpowerup.com/reviews/A...ormance/1.html
These are the results after AMD released their 12.11 driver which was a huge performance boost.
Intel i5-3570K @ 4.7GHz | MSI Z77 Mpower | Noctua NH-D14 | Corsair Vengeance LP White 1.35V 8GB 1600MHz
Gigabyte GTX 670 OC Windforce 3X @ 1372/7604MHz | Corsair Force GT 120GB | Silverstone Fortress FT02 | Corsair VX450
This conversation is strikingly similar to an earlier one, starting here.
the hostility is quite similar too
Milk was a bad choice.
2013 MMO-Champion User of the Year (2nd runner up)
I don't understand why people compare factory OC cards to base their GPU of choice on. Sure a factory OC 670 can be faster than a 680. But that doesn't make the GTX680 slower. Imho, only GHZ edition amd's get the benefit of the doubt, the rest should just be compared stock for performance, and then per brand who has the better OC results (i.e. ASUS DCII, MSI TwinFrozrs). I've had this discussion on multiple techboards and this just doesn't seem to stick..
To be fair, reference GTX 6xx cards don't have much support for overclocking or voltage-tweaking, as NVidia thinks all of their customers are 5 year olds who shouldn't be allowed to do as they please with the big boy toys they buy as adults. ;p
I think you need to compare the cards that are up for question, if it is a reference clocked card or factory overclocked card doesn't matter.
You make comparisons based on the performance the card has when you pull it out of the box, no one can guarantee that you can overclock either card even 5MHz, it might be likely but there just are no guarantees.
Intel i5-3570K @ 4.7GHz | MSI Z77 Mpower | Noctua NH-D14 | Corsair Vengeance LP White 1.35V 8GB 1600MHz
Gigabyte GTX 670 OC Windforce 3X @ 1372/7604MHz | Corsair Force GT 120GB | Silverstone Fortress FT02 | Corsair VX450
Intel i5-3570K @ 4.7GHz | MSI Z77 Mpower | Noctua NH-D14 | Corsair Vengeance LP White 1.35V 8GB 1600MHz
Gigabyte GTX 670 OC Windforce 3X @ 1372/7604MHz | Corsair Force GT 120GB | Silverstone Fortress FT02 | Corsair VX450
I find overclocked cards in relation to the competitors as useless information.
The proper way to measure would be to compare it to other within its family to see how well it overclocks, and then compare the absolute improvement to the opposition and, maybe maybe maybe, perhaps find out how far that family overclocks on average (ie, a sample size and pushing them all to the max) and then compare them to how well the other side overclocks.
7900-series, even the GHz 7970 overclocks better than the GTX 680, and its memory overclocking scales better due to the wider membus.
GTX 600-series respond really well to memory overclocking, which is where they are held back to begin with, its their true 'bottleneck' - memoryspeeds.
That said, both are good cards. I have a GTX 680 and it works fine for what I do with it (watch movies). :P The HD7950 is what really shines though, especially when it comes to overclocking and how far above stock you can go in performance boosts.
Woah, what just happened. Did Tetris say to use overclockability as part of the comparison? Man, I remember back in some HD 7970 discussions I was told to only compare at stock by you guys. :|
I'm not sure what I actually ended up saying (I derail myself in my train of thoughts so commonly I often change sentence mid-sentence, meaning I end being quite confusing).
My intention was to say, it's hard to factor in overclocking since the chips overclock so much differently, and the stock values are either nominal at best, or so differing in attempts to push it between red and green teams (making an example up out of thin air here), like a DirectCuII on the GTX 660 Ti perhaps have a 200MHz overclock wheras the comparable from AMD might just have a DirectCuII HD7950 with a 30MHz overclock.
In short, it's too... flukey. You can compare one cards absolute performance over another cards absolute performance, but then you (general you, not specific on anyone) can't make conclusions like GTX 660 Ti > HD7950, but rather Direct Cu II 660 Ti > Direct Cu II 7950.
While doing this in ever scenario is more accurate, going down to compare them down to this specifics is going to be an exhausting endeavor with too little information posted, unless you just scale up the percentages.
In other words, too much of a headache unless someone's compared two specific cards head-to-head in the same review.
* shrugs