Forum Discussion
Altera_Forum
Honored Contributor
12 years agoSupport for double precision is limited but this is independent of scalar versus vector support. The double precision floating point operators and math functions are not fully conformant with the 1.0 OpenCL spec but you can still use them.