Local Work Size bigger than 128 doesnt work properly

Local Work Size bigger than 128 doesnt work properly

Hi, 

When I try to execute a kernel with a Local Work Size bigger than 128 (1D), for example 256.. the kernel doesnt work properly, giving wrong results. I have the last intel SDK (31360.31441) and I am working on a intel i7 950. I executed the code in NVIDIA and work right....

Suppose that it must works with sizes up to 1024, right? 

Thx in advance!!!!!

2 posts / 0 new
Last post
For more complete information about compiler optimizations, see our Optimization Notice.

Can you attach a small reproducer?

Thanks,
Raghu

Login to leave a comment.