Is it possible to run "for loops" of java code on a GPU using OpenACC ?? my data mining algorithm is encoded in java and I want parallelism my serial code on a GPU. How can I parallelize my Java code on a GPU?
Today, tools are starting to show support for automatic code optimization (Youtube: stochastic optimization for x86 Binaries , Automatic parallelization with Intel Compiler ), but they cannot do this in the general case.
OpenACC is available for Fortran and C ++, no Java support.
This means that today you still have to do it manually yourself. See how the basics of Akka or Hadoop> to make your life easier.
For information on accessing shader languages, see the Java GPU Program .