Nvidia's Pascal GPU architecture will be 10 times faster than Maxwell
At GTC, the GPU Technology Conference, last year, Nvidia announced its new “next-gen” GPU architecture, Pascal. Yesterday, at this year’s conference, NVIDIA CEO and co-founder Jen-Hsun Huang revealed its key features.
Three years of R&D have gone into its development, and there are some impressive numbers that show why Huang was getting so excited about it on stage.
Thanks to NVLink, which links multiple GPUs; improved deep learning through 3D memory; a whopping 32GB of memory, almost 3 times that of the Titan X; and the new mixed-precision computing, it has 10 times the speed of the Maxwell GPUs.
It’s pretty efficient, too, getting 28 teraflops out of 1,300 watts. “If you compare that to a supercomputer in the year 2000, it takes one million watts to deliver one teraflop,” explained Huang.
Pascal GPUs are expected next year.
And I just got the GTX980
I wonder how people plan on cooling 4-GPU graphics cards though... Will water cooling still be enough? Anyway, I guess 2017 will be the year to buy a new graphics card then. Pascal really does look overpowered, sounds like gamer heaven in terms of power available.
Waiting for Pascal is why I'm still running a 670.
I wonder how much more it costs to develop them and how large of a profit margin are they going for...
.
I mean, if they really are 10X as powerful as Maxwell, how much more is it going to cost us?
Because if they can keep the price relatively low, this whole thing sounds like a goddamn GPU revolution....
Oh god, I just picked up a 970 and thought I would be good for a while...
“[Pascal] ten times faster than Maxwell.” Really? Looking closely at the presentation slides, Pascal is _EXTRAPOLATED_ to be 10X faster _IF_the application is bandwidth bound _AND_ it is floating point compute bound _AND_ it can use mixed precision _AND_ it is latency bound by the interconnect, ALL TOGETHER. Those are a LOT of ANDs to happen at the same time… ;-) ...Good marketing piece…