Why do y'all cry gimping when it's obvious Nvidia cards

That's nice, retard.

>>58102318
so why does the 780ti fare worse in games now than a 960, let alone a 3GB 1060?

File: 1476186556995.jpg (16KB, 269x187px) Image search: [Google]

16KB, 269x187px

>>58102368
/thread

Keep gobbling Jen-Hsun Huang's cock OP.

>>58102368
Source? Dx11 or 12?

>>58102410
https://www.youtube.com/watch?v=w_FJyfttrwU

keep in mind this was a card that, at launch, destroyed a 290x.

now it's getting rektd by a 380x.

imo tho, the real reason this is happening is because kepler is a wildly different architecture. if you took a look at the differences between maxwell vs. pascal architecture, they're almost identical so maxwell might not be forgotten, at least in the same way kepler was.

>>58102504
Where's the 960?

>>58102596
During the BF1 bench, go to 4:00

I have thus far only owned AMD products and likely will continue to use them as they've always offered the best bang for buck in my price range (save for 2900 vs 8800 series when they got wrekt up and down) but this gimp meme needs to end. It's as ignorant and stupid as the driver overhead meme.

Nvidia cards have utilized a software based task scheduler for some time now, meaning in order for the card to live up to any kind of potential, it needs competently made drivers for each and every game. Which nvidia is quite good about delivering for whatever their current model range might be, almost always having game ready drivers in advance of or on the day of the release of a new game.
Obviously these guys don't really give a shit about keeping your old card running when they have new stuff out they want to sell. So they focus on the current product line and only provide bug fixes to older cards.
You would know this if you bothered reading up on the new archs when they release.

Meanwhile AMD realized their limited resources and chose to use hardware based task scheduling. It's a brute force approach that allows better performance without competent drivers compared to software based scheduling without competent drivers.
This means they can get away with spending limited resources on getting the big hitting titles optimized, and working on smaller titles at a more relaxed pace.
Couple that with the fact that they've had to carry many of the same cards over for 3 generations, only adding a top performer like the 290x and fury or experimental card like the 285, and you see them in a bit of an odd position.
With no new hardware they shifted focus on keeping those cards at peak potential, which has lead to some pretty considerable gains.

It's been an excellent time to own an amd card, but none of this means they are your best friend nor that the trend is guaranteed to continue.
Both companies are just taking the most logical path to profit.

>>58102368
drivers
everyone knows that amd has worse drivers than nvidia, but slowly over time they work on them to make them better and better.
While nvidia just releases cards with like 95% perfect drivers and can only get that last 5%.

>>58103165
I love my 970

>>58103165
>970 beating 390x/480
TOP KEK
wasn't the 970 "worse" than the 390 (non x)?

>>58103165
The 1060 is only 5 frames off from a fury x?
Consider me fucking sold!

>>58102368
>Projecting

https://www.youtube.com/watch?v=ogMYfj4WFXM

>>58103180
Yes but don't forget that Nvidia on the other side has extra room in overclocking. Maxwell can easily get extra 15% gain on "safe" ocing without touching voltage. Pascal is up there too. Dunno about Vega but latest 480RX had to be cranked like a mofo in order to compete with 1060 gtx which leaves no room what so ever for OC. It's just another tiresome gimmick used by biased AMD reviews when comparing OCing to let's say a 970 gtx - I've noticed it so many times, they will crank that 480 to the point where it's barely safe to use it and adding 5 extra mhz to core clocks would crash it and then they compare it to a factory boost clock of a 970 which could easily gain another 200 mhz on that core and call it a day.


Also thermal throttle should be optional, it pisses me off when the card downclocks itself because it reaches 70 degrees.

>>58103648
>Maxwell can easily get extra 15% gain on "safe"
ha! I wish
my 970 can barely hit 1450mhz over the stock 13something
I blame the dumbass turbo boosting shit, I wish i could turn that off.
Even then I usually get dog shit overclocks on anything I buy.

>>58103668
You either got really unlucky or use some garbage manufacturer. My old asus 970 could easily reach 1530 on core. MSI owners could get it to 1550.