GPUs Set to Increase Performance by 570 Times by 2015 – Chief Executive of Nvidia.

[B.E.E.F.]

Paper is the only thing that matters. After all, that's what the marketing team shows us right?

True, and very embarrassing for these companies. This reminds me of the Phenom launch. Its a paper warrior.

[Cold Fussion]

My 7800gt sli system ran crysis on high at 1920x1200 at less then 1 fps. Now my gtx 260 runs that setup at 30 fps. Thats atleast a 30x jump in 3 years , more considering the 7800gt's were in sli. Computing power is increasing at an exponential rate so its not impossible that it could 500 times faster by 2015.

[highoctane]

Well if you're going to set a goal I guess set it high, really high...

If the trend keeps up with adding more and more parallel cores with process shrinks along with multi gpu scaling who knows what we'll have in another 5+ years compared to today, thats a long ways off yet. It may not translate directly to gaming fps but may be more pointed towards say a Tesla type super computer.

[mstp2009]

And in other news NV predicts they will rename their GPUs 1140 times by 2015.

[firas]

GPUs Set to Increase Performance by 570 Times by 2015

8 times every month?!?!
I’m ok with 30% increase every year, but they can’t even offer that.

[MsB]

Last Night on Coast to Coast, Physicist Richard Phillips that worked on Top Secret Projects and who is a computing specialist said that Moore's law will speed up, not slow down as new technologies are coming into play like nano tech and bio tech computing. He said that it will soon become frustrating to some that as soon as you buy hardware it will be obsolete with next techs being giant steps forward instead of small incremental steps as we are accustomed to. So I believe it. In the future instead of raw speed it will be thousands of nano processors ganged in a group that will do the work instead of just a few large ones like now. Hence cloud computing. He also explained that computers have been held back because man has forced them to compute like the human brain however computers problem solve differently than the human mind and we are learning to take advantage of that.

[Mechromancer]

You know, the more I think about this subject, I doubt Jen Hsun is as crazy as he sounds today. Maybe wer'e being a bit short sighted in the face of all these new material and fabrication technological discoveries as of recent. If we want to make things go faster after 2012 or so, we have to transition to new tech that will allow radically faster tech. We're talking the differences between 8mhz CPU tech, 200mhz CPU tech, and the 3.2Ghz CPU tech today, only that leap will be all at once if we transition from silicon to other materials, like carbon. Nvidia may sound crazy, but they may be on point. I actually hope so. ATI will be right there with them.

[nn_step]

umm moore's law suggest a 16-32 fold increase in performance.

[saaya]

Last Night on Coast to Coast, Physicist Richard Phillips that worked on Top Secret Projects and who is a computing specialist said that Moore's law will speed up, not slow down as new technologies are coming into play like nano tech and bio tech computing. He said that it will soon become frustrating to some that as soon as you buy hardware it will be obsolete with next techs being giant steps forward instead of small incremental steps as we are accustomed to. So I believe it. In the future instead of raw speed it will be thousands of nano processors ganged in a group that will do the work instead of just a few large ones like now. Hence cloud computing. He also explained that computers have been held back because man has forced them to compute like the human brain however computers problem solve differently than the human mind and we are learning to take advantage of that.

mhhh and you believe that? what "expert" was invited in this tv show? he didnt have an asian look to him by any chance? around 40 years old?

what is true is that once optical processing and nano tech take off, we will see "moores law" speed up indeed, like with any new tech, the jump from gen1 to gen2 etc is rather big, until the process/tech becomes wide spread and understood, then things process and advance slower...

but saying computer tech will advance faster with the dawn of nanotech and all is just stupid... sounds like hes either an idi0t or he wanted attention by making sensationalist comments like that...

You know, the more I think about this subject, I doubt Jen Hsun is as crazy as he sounds today. Maybe wer'e being a bit short sighted in the face of all these new material and fabrication technological discoveries as of recent.

right, like this amazing manufacturing breakthrough tsmc achieved with their 40nm process?

umm moore's law suggest a 16-32 fold increase in performance.

moore never said anything about performance