[Various] NVIDIA GeForce GTX 680 Review: Retaking The Performance Crown - Page 52

Originally Posted by Mad Pistol 

2012rig, read the damn graph on Tom's hardware. I don't care what they wrote to downplay their findings. It's black and white and clear as day.
I feel like I'm talking to a bunch of people with their brains turned off. 2 sites have confirmed those findings on power consumption. Nothing else needs to be said.
/done

Originally Posted by 2010rig 

If you say so. The numbers, the graphs, and commentary don't lie.

Originally Posted by XPC 

Actually I don't know what chart you are looking at, but Tom's Hardware's chart clearly shows the GTX680 sitting between the 7970 and 7950, exactly where it's supposed to be. Also, if you actually read THIS page (which is where that chart comes from) and the page that follows it, you will see that Tom's clearly states that the 680 has considerably better performance per watt than the 7970. I think you need to get your eyes checked out, because you are dead wrong on this one.
EDIT: crap 2010rig, you beat me to it!

Originally Posted by Mad Pistol 

I've updated my post. the colors are so close between the HD 6990 and the GTX 680 that it was easy to get them mixed up. The TT review is an anamoly, and based on the power findings of other cards, it really isn't a reliable source anymore. How sad.
There I said it. Enjoy what ever little bit of pleasure you may get out of it.

Originally Posted by dph314 

Anyway... Is this Boost feature going to affect overclocking at all? Will increasing voltage and core clock be the same, just with Boost working/not working depending on how high the user sets the overclock?

Originally Posted by jtom320 


I would expect it with the release of the 720 at this point. It won't be Windows 8 because MS would be talking about it already. Really though no developers are even taking advantage of DX11 yet. We have not had a single game made from the ground up for DX11. I'm a little foggy on this but I don't think we've even had a from the ground up DX10 game so devs still have a long way to go with 11.

Originally Posted by 5entinel 

Folks please always take the TOS and professional initiative into consideration when posting on overclock.net. You are expected to adhere to both. In addition, please drop the stereotype,name calling, posting memes pictures and calling each other "trolls". This is a discussion board.Please respect each other, each other's opinion's and values.
Thank you.

Originally Posted by formula m 




whAaa?


The 720 is already being murmered about, & I have Windows 8 and can run Xbox in VDM.


Secondly, there are many DX11 games out, where have you been...?  <--  dERP!

Originally Posted by XPC 

Well now that all of that power consumption drama is over with, on to some useful conversation.
Does anybody else agree that this is a wonderful direction for Nvidia? Think about it, they have been building massive inefficient, albeit ambitious, dies for their flagships for years now. This time though, they have developed an arch' that is so efficient that a part, which by all of Nvidia's previous design precedents, was intended to be a midrange part performs on the same level (in gaming at least, the jury is still out on compute) with it's competitor's considerably larger and more muscular flagship part. This is a lean chip, and it's a really dense architecture, on the level that we have never seen in the graphics industry.
Now this may not have been an intentional strategy change, with GK100 being scrapped thus necessitating the GK104 flagship upgrade, but hopefully it will show Nvidia that the massive inefficient die method isn't the best way for them anymore. Nvidia has always pushed the limits with their designs, but this time they have channeled that energy into pushing the performance per watt limit. Imagine if instead of the massive GK100 die, Nvidia had worked on a chip that is of similar size to Tahiti. A kepler chip on a die that size would be the indisputable dominant chip this generation, and with it being a smaller chip than GK100, it easily would have been producible enough to bring to market right now.
I hope that with Maxwell Nvidia keeps pushing the efficiency of the architecture further, and then goes for a middle of the road die size for their flagship. It would be nice for them to have a chip that isn't too ambitious to be brought to market on time for a change.