RUS  ENG
Full version
JOURNALS // Program Systems: Theory and Applications // Archive

Program Systems: Theory and Applications, 2014 Volume 5, Issue 4, Pages 3–17 (Mi ps124)

Hardware, software and distributed supercomputer systems

Programming the graphics processors (GPU) in MC# language

Yury Serdyuk

Program Systems Institute of RAS

Abstract: MC# is an extension of the object-oriented language C#. It intended for developing applications running on multicore processors and on clusters with distributed memory. Given paper presents the basic principles and tools for programming the graphics processors (GPUs) in MC# language. The sample program in MC# intended to run on GPU is presented. Further we describe the tools for establishing the parameters of GPU configuration and for using of shared memory in MC# programs. In the final section we outline the current problems under investigation and the directions for further improvement of MC# programming system for GPU. (In Russian).

Key words and phrases: Parallel programming, graphics processors (GPU), shared memory.

UDC: 004.382.2, 004.432

Received: 20.11.2014
Accepted: 09.12.2014



© Steklov Math. Inst. of RAS, 2025