• 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.

Nvidia Kepler vs 500 series specs released - **HOLY CRAP**

Lemme guess, Nvidia have figured out how to bend the laws of Physics yes? Either that or this card is only going to be released in cloud cuckoo land for use by Dipsy, La La and Po.
 
Find this hard to believe myself. A card this good would not be good for the gpu industry as people would not need to upgrade for 6-8 years. Games get slaughtered by the most current hardware already.
 
I'm guessing that the GTR785 is some kind of super overclocked 780, the scaling however just seems fishy. I can't see any of this being genuine unless Nvidia has yet again increased die size. At best these could be theoretical numbers, spat out the day before AMDs hard launch though I'm calling internet BS.
 
This wasn't the one posted some weeks ago. That was 3d mark vantage. Lemme try to find it.

this one...
original.jpg


hmm gains are lower. maybe low vram on 560-590 on current cards
 
Most likely bogus but, its very possible Nvidia can have a few physx effects added, and far beyond me to say Nvidia like to play silly games with physx, but they could easily code it to work worse on Fermi cards than Kepler, making the information right now if it was accurate, useless.

Think of it like this, they add the ability for Kepler to deal with another 2 levels higher tessellation, then they change the physx option in Crysis 2 to use the even higher physx option as default, suddenly it will overload Fermi cards and run fine on Kepler.

When you add stuff you have direct control over the code, the drivers and the hardware...... well, it becomes a laughable comparison.

It was pretty obvious which slides and performance estimates were real for the 7970, largely because everything up till the press day in December was bogus, and everything after looked official and was leaked from multiple sources.
 
going from g80, a 8800gtx to gtx280 was near enough double speed increase.
then going from gtx 280 to gtx 480 was again near enough double speed increase.
its possible kepler will offer double the speed of the gtx480....

(gtx 285, gtx580 were just refreshes of same architecture so not included in above example.)
 
Cyber-Mav said:
going from g80, a 8800gtx to gtx280 was near enough double speed increase.
then going from gtx 280 to gtx 480 was again near enough double speed increase.
its possible kepler will offer double the speed of the gtx480....

(gtx 285, gtx580 were just refreshes of same architecture so not included in above example.)
Click to expand...

It's certainly possible, look at how much faster a 7970 is over a 5870.

Okay, going with those particular product names, the issue isn't how fast the 785gtx is, it's WHEN it will be due. If the 770gtx(which might be more like the 560ti now) is the only thing available up till the end of 2012, and the 8970/780/785gtx come out around the same kind of time..... then that will be the comparison to be made.


The only thing that still makes a lot of sense(the most sense) amongst all the, hot clocks being dropped, mid end coming first, complete change in architecture, same and twice as big, all the differing possibilities, is that Nvidia will have a HUGE core to go much faster than the 580gtx, and that a huge core is unlikely to be available any time soon from Nvidia.

Will be very interesting if that turns out not to be the case but Nvidia has struggled with new processes for a long time now and that isn't helped by huge cores having much lower yields to start off with.
 
You must log in or register to reply here.
Back
Top Bottom