Revision 3fbb12aa928a87a4c4e69a95eaf6f5ee02958600 authored by John Lawson on 09 February 2021, 18:00:02 UTC, committed by GitHub on 09 February 2021, 18:00:02 UTC
* Add support for AVX512 BF16 dot product * Match on f32*f32 * Remove f32 check
1 parent 3e034d6
CodeGen_PTX_Dev.h
#ifndef HALIDE_CODEGEN_PTX_DEV_H
#define HALIDE_CODEGEN_PTX_DEV_H
/** \file
* Defines the code-generator for producing CUDA host code
*/
#include <memory>
namespace Halide {
struct Target;
namespace Internal {
struct CodeGen_GPU_Dev;
std::unique_ptr<CodeGen_GPU_Dev> new_CodeGen_PTX_Dev(const Target &target);
} // namespace Internal
} // namespace Halide
#endif
Computing file changes ...