[Intel-gfx] General Purpose Programming GPU/(G)MCH

[Alan W. Irwin]

On 2010-12-14 10:42+0100 Clemens Eisserer wrote:

>>  .. they start call gpu inside GMCH since 965 ). I wold be great full for some more links to decumentation and
>> codes.
>
> I guess just because the term "GPU" was uncommon before. Fact is, gen3
> contains programmable pixel shaders.
> However if it isn't just for fun I discourage from going that route,
> the shaders are extremly slow compared to todays CPUs (keep in mind
> gen3 is now ~5 years old).
>

I personally wouldn't be quite that discouraging on the speed front.
Today's cpu's are of similar clock speed to those of yesteryear.  I bought
an entry level 2.4GHz box back in 2003, and entry level boxes are
still of that order in raw speed.  The real advances for modern cpu's
are computing power efficiency (gigaflops/watt), and number of cpu's.
Thus, modern cpu's give you a real speed advance only if your
application can take advantage of multiple cpu's, but that is often
not the case. I agree some speed advances have been made in the Intel
GPU case (see http://en.wikipedia.org/wiki/Intel_GMA for some of the
speed details), but as in the CPU case it again appears to me that
most of the advances in the GPU case are in the number of execution
units.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________