gpu architecture name

[quadboon]

afaik there is currently no option in opencl to “cross compile” to other gpu architectures. in cuda this is possible by adding the --arch option to nvcc. but please correct me if i am wrong. i really need this option since my app is distributed with binary kernels.