Fix bitcasts with clang on x86 bitcasts.h fails to compile with clang on x86. clang supports x86 Intel intrinsics (such as _castu32_f32) but the appropriate header must be included These intrinsics are defined in ia32intrin.h [1] which must be included via x86intrin.h [2] [1] https://clang.llvm.org/doxygen/ia32intrin_8h.html#a856b4d4dc3e7300cce3277ac12368687 [2] https://clang.llvm.org/doxygen/ia32intrin_8h_source.html
Header-only library for conversion to/from half-precision floating point formats
The library is developed by Marat Dukhan of Georgia Tech. FP16 is a research project at Richard Vuduc's HPC Garage lab in the Georgia Institute of Technology, College of Computing, School of Computational Science and Engineering.
This material is based upon work supported by the U.S. National Science Foundation (NSF) Award Number 1339745. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect those of NSF.