Sign in with
Sign up | Sign in
Your question

Video card computing question

Tags:
  • Graphics Cards
  • Gtx
  • Games
  • Power
  • Graphics
Last response: in Graphics & Displays
Share
May 30, 2012 1:36:35 AM

I have read often recently that some newer Nvidia cards such as the GTX 680 have reduced computing power, but that it's great for games.

What the heck exactly does "computing" power mean in this case, and for what kind of processes is this so-called computing power used for in a video card? GTX 680 is great for games, but what isn't it great for because of reduced computing power?

More about : video card computing question

a c 291 U Graphics card
May 30, 2012 6:11:05 AM

Computing is calculating. There has been a trend lately to use graphics cards for parallel calculations, because they are much faster than CPUs at it.

Quote:
The reason behind the discrepancy in floating-point capability between the CPU and the GPU is that the GPU is specialized for compute-intensive, highly parallel computation – exactly what graphics rendering is about – and therefore designed such that more transistors are devoted to data processing rather than data caching and flow control.

More specifically, the GPU is especially well-suited to address problems that can be expressed as data-parallel computations – the same program is executed on many data elements in parallel – with high arithmetic intensity – the ratio of arithmetic operations to memory operations. Because the same program is executed for each data element, there is a lower requirement for sophisticated flow control, and because it is executed on many data elements and has high arithmetic intensity, the memory access latency can be hidden with calculations instead of big data caches.

Data-parallel processing maps data elements to parallel processing threads. Many applications that process large data sets can use a data-parallel programming model to speed up the computations. In 3D rendering, large sets of pixels and vertices are mapped to parallel threads. Similarly, image and media processing applications such as post-processing of rendered images, video encoding and decoding, image scaling, stereo vision, and pattern recognition can map image blocks and pixels to parallel processing threads. In fact, many algorithms outside the field of image rendering and processing are accelerated by data-parallel processing, from general signal processing or physics simulation to computational finance or computational biology.


You can read more here: http://developer.download.nvidia.com/compute/DevZone/do...
m
0
l
June 6, 2012 1:51:29 PM

Sunius said:
Computing is calculating. There has been a trend lately to use graphics cards for parallel calculations, because they are much faster than CPUs at it.

Quote:
The reason behind the discrepancy in floating-point capability between the CPU and the GPU is that the GPU is specialized for compute-intensive, highly parallel computation – exactly what graphics rendering is about – and therefore designed such that more transistors are devoted to data processing rather than data caching and flow control.

More specifically, the GPU is especially well-suited to address problems that can be expressed as data-parallel computations – the same program is executed on many data elements in parallel – with high arithmetic intensity – the ratio of arithmetic operations to memory operations. Because the same program is executed for each data element, there is a lower requirement for sophisticated flow control, and because it is executed on many data elements and has high arithmetic intensity, the memory access latency can be hidden with calculations instead of big data caches.

Data-parallel processing maps data elements to parallel processing threads. Many applications that process large data sets can use a data-parallel programming model to speed up the computations. In 3D rendering, large sets of pixels and vertices are mapped to parallel threads. Similarly, image and media processing applications such as post-processing of rendered images, video encoding and decoding, image scaling, stereo vision, and pattern recognition can map image blocks and pixels to parallel processing threads. In fact, many algorithms outside the field of image rendering and processing are accelerated by data-parallel processing, from general signal processing or physics simulation to computational finance or computational biology.




You can read more here: http://developer.download.nvidia.com/compute/DevZone/do...


I've got all the info now, thanks.

To boil it down a little, for someone like myself who uses Luxology modo, The Foundry Mari and NextLimit Maxwell Render for 3D content-creation (push 3D polygons around, paint, render), how should I decide between GTX 680 4GB and GTX 580 3GB. Note, The Foundry Mari lists GTX 580 as an approved/tested card and their testing team have given the nod to GTX 680, though not yet listed on their site. They would have been testing GTX 680 2GB, not the 4GB version just now hitting the shelves. Higher VRAM is important for a smooth painting and sculpting experience in modo and Mari, btw. Either card will no doubt be fine for the polygon modeling in modo. Don't know about render.

In any case, again, between GTX 580 3GB and GTX 680 4GB for 3D work at the serious enthusiast level, I'm in a quandary. I will not be considering the Nvidia Quadro line or ATI products. I've read so much over the past few months, I think I'm back to where I don't know a thing :)  . My botttom line question, I guess, is this: is the GTX 680 4GB "held back" by Nvidia to the poiint where it is actually a less desirable card overall for my kind of work than the GTX 580 3GB? This exact question has been posted elsewhere in numerous forums regarding 3D work but hasn't actually been answered in a direct Yes/No fashion, as far as I know. The info I read is always in regards gaming, in which I am not interested.

Thanks.

m
0
l
!