Re: Mathematica 8
- To: mathgroup at smc.vnet.net
- Subject: [mg113923] Re: Mathematica 8
- From: Murray Eisenberg <murray at math.umass.edu>
- Date: Thu, 18 Nov 2010 07:05:05 -0500 (EST)
If that response cited AMD, then I apologize for missing it. If not, then it's no wonder I didn't see the connection to AMD graphics cards. In any event, the support for OpenCL, as implemented at the GPU end by AMD graphics cards, seems to be getting a decidedly secondary -- indeed, perhaps, nearly hidden -- mention in the publicity for Mathematica 8. Thus in the "News" area overlaying http://wolfram.com, one sees "Pioneering Free-Form Linguistic Input"; "Revolutionizing ePublishing", and "Simplifying CUDA Programming". And the last-mentioned itemsays nothing about AMD or OpenCL and has a link to a wihte paper on . If you look at the page http://www.wolfram.com/mathematica/new-in-8/, then you do see a reference to CUDA and OpenCL finally. And on the target of that link, finally reference to AMD, too. I'd like to see some discussion in this group about advantages/disadvantages of CUDA vs. OpenCL, especially when used in conjunction with Mathematica 8? I'll have to upgrade my hardware from an ATI board that's too old to support OpenCL, so I want to choose wisely between ATI and nVidia, with now Mathematica 8 in the picture. On 11/17/2010 9:10 AM, John Fultz wrote: > I also indicated (just a few words after the part of my response you quoted) > that v8 supports OpenCL. Many AMD cards support OpenCL. > > Sincerely, > > John Fultz > jfultz at wolfram.com > User Interface Group > Wolfram Research, Inc. > > > On Wed, 17 Nov 2010 05:31:17 -0500 (EST), Murray Eisenberg wrote: >> Does ATI/AMD offer anything like CUDA that might be the target of future >> Mathematica development? >> >> I ask because I'd hate to think I'd be condemned to using ATI/AMD >> graphics cards to take advantage of the GPU from Mathematica. I say this >> since my experience over many years is that nVidia graphics drivers have >> been less stable than ATI drivers. (But that's just my own experience, >> and of course others' may vary.) >> >> On 11/16/2010 5:05 AM, John Fultz wrote: >> >>> Other new features related to this functionality include support for >>> CUDA > > -- Murray Eisenberg murray at math.umass.edu Mathematics & Statistics Dept. Lederle Graduate Research Tower phone 413 549-1020 (H) University of Massachusetts 413 545-2859 (W) 710 North Pleasant Street fax 413 545-1801 Amherst, MA 01003-9305