openCL

OpenCL & CUDA post 10.6.4

cybero's picture

Just to let you all know there has already been a bug filed for the way 10.6.4 knocks the previously accepted CUDA driver of its perch.

Octane, for instance, doesn't launch.

An explicit message regarding this is to be found in the CUDA preferences panel, somewhat self contradictory too as it at one and the same time states that the software is up to date and yet that the driver needs an update.

Thus far it seems to have made little or no difference to anything else OpenCL wise, but CUDA is proving to be a different matter, though that isn't essentially relevant to this forum.

OpenCl Open Gl Interop

mattgolsen's picture

I thought you guys might find this interesting, stumbled across it tonight

The article is for C#, but pretty cool nonetheless.

http://www.cmsoft.com.br/index.php?option=com_content&view=category&layo...

Snow.qtz Javascript Fix and Particle Count Mod

gtoledo3's picture

When looking at a recent post by yanomano about array size, I realized that the Snow.qtz example from the Apple Developer Examples "OpenCL" download suffered from a very similar code setup to set the particle positions.

I have changed that so that when the particle count is reduced, that the structure reduces correctly.

Also, I have noted that when the qtz is initialized, that it runs a bit "oddly" until particle count is reset. This amounts to the distribution of particles not being as even looking, and performance being a bit more uneven. I'm not sure if this has always been the case with this composition, but I've added a small portion to the setup to create a pulse at startup that quickly adds and subtracts a particle on startup, and any time that particle count is changed thereafter. This smooths performance on my system.

OpenCL : Array Size at run time

yanomano's picture

Is there a way to force an OpenCL kernel to adapt its working size if an array size change on its input? the function get_global_size() seems to evaluate only one time at the start.

OpenCL Image Processing Routines

cybero's picture

Please find attached an example file that exhibits the use of six OpenCL image processing routines.

They have all been ported over from noise_kernels.cl example from Apple's developer code examples.

They will happily take a static image or a moving one.

The posted example uses a static image, the online screen grab posted to vimeo uses a video feed [some HD footage of my own] and the output of the kernels is used there as a mask image to the main image, which is also the product of a CoreImage processing routine that inverts the source image .