Clik here to view.
Re: Maths in double precision gives me different results on AMD and NVIDIA
Hi yurtesen,help me to understand why there is a race condition if the global_size and the local_size are 1?Thanks. Bwt using this modified kernel gives same results: __kernel void div_kernel2(...
View ArticleRe: Maths in double precision gives me different results on AMD and NVIDIA
Yes that's what i pointed out in my reply.
View ArticleRe: How to use GpuPerfStudio 3 to get gpu info from Chrome running a WebGL...
Thanks for the info.--no-sandbox on the command line does the trick. 2014-10-25 1:44 GMT+02:00 tonyhosier <developer.forums@amd.com>: AMD Developer...
View ArticleClik here to view.
Performance issues with glBufferSubData.
Hi! I have recently implemented a system which allows our game engine to use persistent and coherently mapped buffers using OpenGL. While this gives a massive speed boost when it comes to mapping...
View ArticleRe: Maths in double precision gives me different results on AMD and NVIDIA
musale87 wrote: The solution (suggested by an user of GPGPU group on linkedIn) was to made my own pow() function like this: double my_pow(double x, double a) { return exp(x * log(a)); } It...
View ArticleHow to process images using SVM of OpenCL2.0?
As per the (section 5.6.1 from OpenCL 2.0 spec, page 166),we try to create Image objects using SVM...
View ArticleRe: Opengl interop / clCreateContextFromType segfault
I found the problem ... or at least a workaround: Obviously the wrong libGL was used for the case that OpenGL interoperability is made use of. I changed the following symlink:...
View ArticleRe: Re: Maths in double precision gives me different results on AMD and NVIDIA
Well sorry, it was my mistake, I didn't check your code attachment, so I didnt realize you enqueue the kernels 1000 times etc. Now I downloaded your kernel and ran it on Spectre, Hawaii, AMD A10-7850K...
View ArticleRe: Re: Re: Maths in double precision gives me different results on AMD and...
Thanks yurtesen! your result is good!I'm seriously thinking that I have a problem in configuration or hw, may be a problem with the driver which goes in conflict with nvidia ones or problem with power...
View ArticleRe: How to process images using SVM of OpenCL2.0?
Hi As per OpenCL 2.0 spec:Although SVM is generally not supported for image objects, clCreateImage may create animage from a buffer (a 1D image from a buffer or a 2D image from buffer) if the buffer...
View ArticleClik here to view.
Re: How to get the status of multi-monitor by W600 video card / SDK?
Hi,you can have a look at the ADL SDK: http://developer.amd.com/tools-and-sdks/graphics-development/display-library-adl-sdk/ ADL gives you the possibility to get some state information on the...
View ArticleRe: OpenCL™ 2.0 Preview
Thanks for the nice blog post with a clear intro to coarse SVM. I like the simplicity of the pointers being shared but am less sure about the performance implications in the example presented. Is there...
View ArticleRe: Maths in double precision gives me different results on AMD and NVIDIA
Thank you very much! I will install the operating system from scratch and I will try the system without the nvidia card I hope that all goes well I'm sorry for this wrong discussion
View ArticleRe: List of platforms supported by OpenCL 2.0 driver
MIchael, I have some answers I can share. Will AMD be supporting Fine Grain Shared Virtual Memory on existing or future graphics cards (perhaps only on AMD based motherboards) This is not a core...
View ArticleRe: Exception calling clBuildProgram on my kernel
Dipak, Thanks, but I really don't understand this kind of response from you.An exception when compiling is clearly a bug in the AMD compiler.And I have given you a reproducer. So, please try to...
View ArticleRe: OpenCL 2.0 example compile problem
Finally, the following command works for me. # ln -s /usr/lib/libOpenCL.so.1 /usr/lib/libOpenCL.so
View ArticleRe: OpenCL 2.0 example compile problem
Hi,Thanks for reporting this.There are two mode of Catalyst installation: 1) Install Driver <Os version> on X.org 6.9 or later and 2) Generate Distribution Specific Driver Package (Default). I...
View ArticleVery slow drawing / filling of ellipses (GDI+ and Direct2D) on YUV422 surfaces.
Using GDI+, the Graphics.DrawEllipse and Graphics.FillEllipse functions are currently taking ~100ms using the latest Catalyst driver in Vista and newer operating systems. These same function calls...
View ArticleRe: hsa kernel driver and radeon module
>>>>As bsp2020 said, we have not yet released an OpenCL runtime which works over the HSA stack. Short term focus is on enabling new languages & interfaces, eg. Java, C++ AMP, OpenMP...
View Article