I think the "-d16" suffix means it has 16 double precision floating point registers, which equates to 32 single precision registers.
Statistics: Posted by alastairpatrick — Sun Sep 08, 2024 9:27 pm