Halide  17.0.2
Halide compiler and libraries
EmulateFloat16Math.h File Reference

Methods for dealing with float16 arithmetic using float32 math, by casting back and forth with bit tricks. More...

#include "IR.h"

Go to the source code of this file.

Namespaces

 Halide
 This file defines the class FunctionDAG, which is our representation of a Halide pipeline, and contains methods to using Halide's bounds tools to query properties of it.
 
 Halide::Internal
 

Functions

bool Halide::Internal::is_float16_transcendental (const Call *)
 Check if a call is a float16 transcendental (e.g. More...
 
Expr Halide::Internal::lower_float16_transcendental_to_float32_equivalent (const Call *)
 Implement a float16 transcendental using the float32 equivalent. More...
 
Expr Halide::Internal::float32_to_bfloat16 (Expr e)
 Cast to/from float and bfloat using bitwise math. More...
 
Expr Halide::Internal::float32_to_float16 (Expr e)
 Cast to/from float and bfloat using bitwise math. More...
 
Expr Halide::Internal::float16_to_float32 (Expr e)
 Cast to/from float and bfloat using bitwise math. More...
 
Expr Halide::Internal::bfloat16_to_float32 (Expr e)
 Cast to/from float and bfloat using bitwise math. More...
 
Expr Halide::Internal::lower_float16_cast (const Cast *op)
 Cast to/from float and bfloat using bitwise math. More...
 

Detailed Description

Methods for dealing with float16 arithmetic using float32 math, by casting back and forth with bit tricks.

Definition in file EmulateFloat16Math.h.