Riya Bisht

Google Summer Of Code 2024 Contributor

email: manasi.riya2003@gmail.com

Education: B.Tech in Computer Science Engineering, Graphic Era University, India

Ongoing project: Enable CUDA compilation on Cppyy-Numba generated IR
Cppyy is an automatic, run-time, Python-C++ binding generator, for calling C++ from Python and Python from C++. Initial support has been added that allows Cppyy to hook into the high-performance Python compiler, Numba which compiles looped code containing C++ objects/methods/functions defined via Cppyy into fast machine code. The project aims to enhance Cppyy, by enabling CUDA compilation on Numba-generated intermediate representation (IR). This integration will allow seamless utilization of CUDA paradigms in Python without compromising performance. The objectives includes adding support for Cppyy-defined CUDA Code i.e. implementing support for declaration and parsing of CUDA code defined in Cppyy within the Numba extension. Designing and developing a CUDA compilation mechanism and execution within the Cppyy-Numba environment.

Project Proposal: URL

Mentors: Vassil Vassilev, Wim Lavrijsen, Aaron Jomy, Jonas Rembser

Presentations



Enable CUDA Compilation on Cppyy-Numba generated IR, Slides, Team Meeting, 29 May 2024