Clik here to view.
Instruction throughput clarification
Hi,I am reading AMD_Accelerated_Parallel_Processing_OpenCL_Programming_Guide-rev-2.7.pdf and on page 6-24 (or 126 in absolute terms) there is: Table 6.3 Instruction Throughput (Operations/Cycle for...
View Articlepassing SVM pointers in struct?
Can you pass SVM pointers to kernel in a struct by value? I'm trying to do this, but I get a segmentation fault when I set the argument with AMD's drivers. It works with intel's opencl drivers, though....
View ArticleOpenCL™ 2.0 Demystified – Device-side Enqueue
Our blogs on SVM and Pipes showed you how OpenCL 2.0 can make your applications simpler. But OpenCL 2.0 isn’t only about programmability. It’s also about performance. Our new blog post explains how...
View ArticleClik here to view.
RLEJPEG-Sample-Files dosn't work on my PC which has AMD graphic card.
hello. i'm beginner of develope AMD codes. So, I downloaded sample files on AMD's Developer site, which is RLDJPEG-Sample-Files. This source codes is build success and can get result of codes in my...
View Articlesegmentation fault inside clBuildProgram (bug demonstration attached)
I have a particular OpenCL program that is revealing a bug with clBuildProgam for AMD CPU device. After several recent code changes, my OpenCL kernel/program will compile just fine on Apple and NVIDIA...
View ArticleList of platforms supported by OpenCL 2.0 driver
We have updated the list of AMD products supported by the early release of the OpenCL 2.0 driver. You can find that list here: The original page listed supported product families, but in some cases...
View ArticleclSetKernelArgSVMPointer() does not support a constant?
main.cpp: cl_uchar4* svmbuffer=(cl_uchar4*)clSVMAlloc(....)cl_uchar4* svmoutbuffer=(cl_uchar4*)clSVMAlloc(....)int n=2;...............status = clSetKernelArgSVMPointer(kernel2D,0,svmbuffer);status =...
View Article14.9 and higher OpenCL performance
I've tried dozens of tweaks to my Groestl implementation, and it refuses to get anywhere near as fast as pre-14.9 code. The pre-14.9 code is just butchered on 14.9 - a rewrite makes it better, but the...
View ArticleSharing my GPU setup with public
Hi all,In my honest opinion, most High-end GPGPU capable machines sit idle for a considerable time. I recently transformed my old laptop, hosting a HD5870M into a GPGPU server, which can be accessed by...
View ArticleClik here to view.
Suggest Feature you want in AMD APP
Suggest Feature you want in AMD APPHi EveryBody,I was preparing a top feature requests for the AMD's openCL implementation. I will be looking to as many old forum topics as i can. But there can always...
View Articlesegfault at clBuildProgram
Hi, after modifying a number of variable declarations fromdouble to single precision in a (working) .cl file, I now have a systematic segmentation fault at the clBuildProgram call.It does not seem to...
View ArticleHow to use OpenCL via SSH for normal users?
I've been trying to find some way to let remote users run OpenCL programs via SSH without them having to start up a local X session. The closest I've found is here but it still requires an X session...
View ArticleWhat's the type of SVM cache supported?
SVM logically extends global memory into host address space, and the cache support for global memory on Kaveri is CL_READ_WRITE_CACHE.Does it mean that SVM will be cached as read and write type? Does...
View ArticleRelational built-in function 'select' causes "Segmentation fault"
Hi, I'm trying to use the select built-in function in the following kernel : __kernel void Kernel1(__global uchar4 *a, __global uchar4 *b, __global uchar4 *c,...
View Articleglobal memory write cache race condition under load - Opencl 1.1, 14.9 and...
Hi, I have the following sync issue:I have a kernel that does some output to a buffer. after calling the kernel I issue a cl::finish (also tried waitforevents, enqueueBarrier etc.) on that queue and...
View ArticleStatus of OpenCL 2.0 SDK and reference manual?
Initially the updated APP SDK was expected this summer. Is this still the case? Current APP programming guide is too outdated! No information is included about Bonaire and later GPUs and Sea islands or...
View ArticleGuide to run OpenCL headless, without X server and as normal user
You may know that you can run OpenCL program as root without Xserver. I found patch which enable running the OpenCL programs even without Download catalyst driver. Then run chmod +x...
View ArticleGlobal to private memory copy
Hello, I have a kernel where each instance of the kernel must access 16 floats from global memory (in series) and perform some computation on them. Adjacent kernels may or may not need to access the...
View ArticleAPU + Discrete Card
Hello, Will it be possible to have an AMD APU (such as upcoming Carrizo) and and AMD discrete card (such as the upcoming 300 series)on one system and be able to use both discrete GPU and integrated...
View Article