lowering the use of the memory controller in OpenCL based applications

Posted by user827992 on Programmers See other posts from Programmers or by user827992
Published on 2012-07-03T15:51:19Z Indexed on 2012/07/03 21:22 UTC
Read the original article Hit count: 307

Filed under:

opencl

With my first experiments I noticed that OpenCL is a good technology but often hampered by the X86 architecture and finding a mid-range VGA driven by a low-end chipset is not that unusual in the real world scenarios, sometimes this can happen with some high-end VGA too.

Are there some caching techniques? Something that can bypass this inconvenience in some ways. The amount of dedicated memory on today's VGA is usually high, it's possible to use this memory to create some kind of buffer with instructions.

Related posts about opencl

QtOpenCl make errors. Please help.

as seen on Stack Overflow - Search for 'Stack Overflow'
So I downloaded the ATI Stream SDK. I don't have a gpu now so I use the '-device cpu' and got the programs/examples in the OpenCl directory working by adding the directory to LD_LIBRARY_PATH etc. Now the problem is when installing QtOpenCl. configure script gives me: skkard@skkard-desktop:~/Applications/qt-labs-opencl$… >>> More
How do I use local memory in OpenCL?

as seen on Stack Overflow - Search for 'Stack Overflow'
I've been playing with OpenCL recently, and I'm able to write simple kernels that use only global memory. Now I'd like to start using local memory, but I can't seem to figure out how to use get_local_size() and get_local_id() to compute one "chunk" of output at a time. For example, let's say I wanted… >>> More
Linux QT OpenCL basic setup

as seen on Stack Overflow - Search for 'Stack Overflow'
Hi, what's the basic setup for Linux to compilie a C/C++ examples from OpenCL SDK? >>> More
solve a classic map-reduce problem with opencl?

as seen on Stack Overflow - Search for 'Stack Overflow'
I am trying to parallel a classic map-reduce problem (which can parallel well with MPI) with OpenCL, namely, the AMD implementation. But the result bothers me. Let me brief about the problem first. There are two type of data that flow into the system: the feature set (30 parameters for each) and… >>> More
Custom types in OpenCL kernel

as seen on Stack Overflow - Search for 'Stack Overflow'
Is it possible to use custom types in OpenCL kernel like gmp types (mpz_t, mpq_t, …) ? To have something like that (this kernel doesn't build just because of #include <gmp.h>) : #include <gmp.h> __kernel square( __global mpz_t* input, __global mpz_t number, __global int* output… >>> More

Developer IT