Attending this event?
View analytic
Friday, September 28 • 09:00 - 10:00
Bringing C++ 17 Parallel Algorithms to a standard library near you

Sign up or log in to save this to your schedule and see who's attending!

There are two proposed parts of this talk -
Part 1 - Integration:

Intel have offered to contribute an implementation of the C++17 parallel algorithms to libstdc++ (GCC)
and libc++ (Clang). This talk will cover -
* brief introduction to C++17 parallel algorithms and execution policies
* overview of Intel's implementation
* considerations and changes necessary to integrate a substantial 3rd party contribution like this
into a standard library implementation.
* support for different execution back ends, tailored to the target platform

Part 2 - Mapping parallel algorithms to the proposed Executors design:

There is strong support for Executors to be part of the next version of C++. Support for parallel algorithms targeting
the proposed Executors design will be an important part of that work. The Intel parallel algorithms implementation
currently being integrated into libstdc++ and libc++ provides basic support for pluggable execution environments.
This talk would cover integrating the current Executors design as a concrete execution target for the parallel
algorithms implementation, with a focus on how you would use the proposed Executors API as a target for common
parallel algorithm implementations.


Thomas Rodgers

Sr Software Engineer - Platform Languages Runtime, RedHat, Inc
Thomas has been programming in C++ since 1989 and a member of the C++ Standards Committee since 2013. Thomas has worked in the financial industry since 1996 and currently works for DRW Trading Group in Chicago.

Friday September 28, 2018 09:00 - 10:00

Attendees (25)