Intel® IPP 2019 Update 3 is now available. Intel IPP 2019 packages are now ready for download. Intel IPP is available as part of the Intel® Parallel Studio XE and Intel® System Studio. Please visit the https://software.intel.com/en-us/intel-ipp
What's New in Intel® IPP 2019 Update 3:
- Extended Intel® AVX2 and Intel® AVX-512 optimization for ippsFIRMR32f_32fc functions.
- Added a threading layer example on building custom pipeline code with the sobel filter functions.
- Added new implementation in the IIR Filter functions to improve precision. The implementation provides similar performance for long length vectors, but reduces the performance for short length vectors.
- Fixed several bugs in the bzip2 optimization patch files, including Makefile errors, and inconsistency with original code.
- Fixed the following dispatching issues in Intel® IPP libraries:
- The IPP functions dispatched Intel® AVX2 optimization code on the Intel® AVX-512 processors with macOS* systems.
- The cryptography libraries dispatched incorrect code on the systems that do not support Intel® AVX instructions.
- The cryptography libraries dispatched incorrect optimization code when the libraries were linked dynamically.
Check Intel® IPP 2019 release notes to learn more information.