In response to Marcus: Of course they work that way. 

But they are far from monolithic. For example, Nvidia was the company that 
pushed OpenGL 3.0 to announce that starting with 3.1, a tremendous amount of 
functionality would be deprecated, thus rendering most existing OpenGL 
applications to the trash bin. Then other Nvidia people realized that was a 
terrible position pushed  Open GL 3.2 and up to support both CORE and 
COMPATIBILITY profiles.

In response to Russell: Cg was light years ahead of GLSL but GLSL overtook it 
and even though there are elements of Cg that are better than GLSL. GLSL is 
core to OpenGL and has evolved so it performs well. Cg is now pretty head. I 
would expect the same to happen with CUDA and OpenCL. What may be more relevant 
is WebGL and the soon to be released WebCL.

In practice, the standards are less designed by committee than by a few members 
of the committees who put in infinite time. In many cases, this winds up to be 
far worse than design by committee. Most of the recent graphics standards come 
from the work of a few driver writers who have almost no contact with the 
application community. I've had discussions with Kronos at various meetings 
about the problem of their committees being dominated by a handful of large 
commercial players. Their attitude is that the research and educational 
communities can participate if they buy memberships (and that would imply 
having someone with the time to spend).

Mesa is a viable alternative except that you are still stuck with the standard 
that comes out of Kronos.

Ed
__________

Ed Angel

Founding Director, Art, Research, Technology and Science Laboratory (ARTS Lab)
Professor Emeritus of Computer Science, University of New Mexico

1017 Sierra Pinon
Santa Fe, NM 87501
505-984-0136 (home)                     [email protected]
505-453-4944 (cell)                             http://www.cs.unm.edu/~angel


On Feb 8, 2013, at 4:22 PM, Russell Standish wrote:

> Added to that, OpenCL was light years behind CUDA (at least that was
> the case two years ago, when I looked at it last). I can understand
> nVidia making sure their product is OpenCL compatible, but putting
> their R&D into CUDA. To be quite honest, it is damn hard to get
> cutting edge R&D into APIs designed by a committee, and that probably
> the way it should be.
> 
> More reprehensible is their attitude towards the Open Source Nouveau
> driver, although that may have improved since Linus spat the dummy at
> them.
> 
> 
> On Fri, Feb 08, 2013 at 05:45:46PM -0500, [email protected] wrote:
>> I wrote:
>> 
>> "For example, for a while AMD used OpenCL to discriminate
>> themselves from nVidia (the little guy in *that* example)."
>> 
>> To clarify,
>> 
>> nVidia likes OpenGL -> Microsoft is competitor with DirectX.  Competitors
>> are bad.
>> AMD likes OpenCL -> nVidia competitor with CUDA.  Competitors are bad.
>> nVidia doesn't invest OpenCL -> No serious competitor to CUDA
>> 
>> [etc]
>> 
>> --------------------------------------------------------------------
>> mail2web.com - Microsoft® Exchange solutions from a leading provider -
>> http://link.mail2web.com/Business/Exchange
>> 
>> 
>> 
>> ============================================================
>> FRIAM Applied Complexity Group listserv
>> Meets Fridays 9a-11:30 at cafe at St. John's College
>> to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> 
> -- 
> 
> ----------------------------------------------------------------------------
> Prof Russell Standish                  Phone 0425 253119 (mobile)
> Principal, High Performance Coders
> Visiting Professor of Mathematics      [email protected]
> University of New South Wales          http://www.hpcoders.com.au
> ----------------------------------------------------------------------------
> 
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com

Reply via email to