Why can’t I use a zero increment in cublasSetVector()?
A zero increment is valid in cublasDCopy, as well as mkl etc., and it’s very handy.
How can this be requested for a future cublas releases?
Thanks
Feature requests should be filed through the bug reporting form linked from the registered CUDA developer website. Please prefix the synopsis (subject line) of the bug report with “RFE:” (short for “request for enhancement”).