Tracy Harms wrote:
With regard to the possibilities regarding parallelism
and J, I posted to Jchat on Sat Dec 30 00:59:01 HKT
2006.  My message is entitled 'parallel processing via
NVIDIA CUDA'.  In contrast with the "coarse-grain
parallelism" available with dual-core and quad-core
main processors, fine-grained hardware innovations
such as this one provide array-like processor
configurations.  The potential for this to fit with
array programming languages such as J seems natural.


Does this NVIDIA support IEEE double precision floating point?

--
regards,
bill
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to