accelerator::get_supports_double_precision Method
TOC
Collapse the table of content
Expand the table of content

accelerator::get_supports_double_precision Method

 

Returns a Boolean value that indicates whether the accelerator supports double precision math, including fused multiply add (FMA), division, reciprocal, and casting between int and double.

bool get_supports_double_precision() const;

true if the accelerator supports double precision math; otherwise, false.

Requirements

Header: amprt.h

Namespace: Concurrency

Show:
© 2016 Microsoft