jcuda or cuda: pros and cons of over versus the other [closed]

Question

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.

Closed 9 years ago.

Me mother tongue is Java (I know I know, but you can't change who you are).

So I wanted to get a pros / cons of doing CUDA vs. JCUDA.

I just found out there is a jcuda.org

Thanks in advance.

Answer 1

cuda represents language extensions to C or C++ (or Fortran). JCuda presents a binding API so that library calls (e.g. cublas for linear algebra, or cufft for fft's, etc.) and arbitrary cuda functions written in C or C++ can be called from Java. If you want to write java kernel code that runs on the GPU, you may also be interested in root beer. Root beer allows you to write kernel code that runs on the thread processors in a GPU directly, without any requirement for CUDA C. However, you'll still need to install the CUDA toolkit to enable the Root Beer "Java Compiler".