[automerger skipped] Empty merge of Android 24Q2 Release (ab/11526283) to aosp-main-future am: e0e33d6a50 -s ours

am skip reason: Merged-In I0be9ba41fc16b264a01fd3bc94a06cf51a3270a8 with SHA-1 1ded9b2880 is already in history

Original change: https://googleplex-android-review.googlesource.com/c/platform/external/FP16/+/27144227

Change-Id: I2323caf78fc54a3bf26a6dbeffc64c012d8b12bf
Signed-off-by: Automerger Merge Worker <[email protected]>
tree: cb376e49da4efb0db30089e589c64672c41dd208
  1. bench/
  2. cmake/
  3. include/
  4. test/
  5. .gitignore
  6. Android.bp
  7. CMakeLists.txt
  8. configure.py
  9. confu.yaml
  10. LICENSE
  11. METADATA
  12. MODULE_LICENSE_MIT
  13. OWNERS
  14. README.md
  15. TEST_MAPPING
README.md

FP16

Header-only library for conversion to/from half-precision floating point formats

Features

  • Supports IEEE and ARM alternative half-precision floating-point format
    • Property converts infinities and NaNs
    • Properly converts denormal numbers, even on systems without denormal support
  • Header-only library, no installation or build required
  • Compatible with C99 and C++11
  • Fully covered with unit tests and microbenchmarks

Acknowledgements

HPC Garage logo Georgia Tech College of Computing logo

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.