Multiprocessor Machines and Parallelization of Algorithms (Applied Mathematics)

Type: For the student's choice

Department: computational mathematics

Curriculum

SemesterCreditsReporting
94Exam

Lectures

SemesterAmount of hoursLecturerGroup(s)
918

Laboratory works

SemesterAmount of hoursGroupTeacher(s)
936

Course description

Aim. This course is devoted to the construction of parallel forms and systolic arrays computational algorithms and parallelization of numerical methods of linear algebra.

Summary. The course covers the main points of operation multiprocessor machines, construction of parallel forms of algorithms and systolic structures parallelization algorithms for basic numerical linear algebra methods.

Target. Read, explore various forms of parallel algorithms and systolic arrays and be able to parallelize the numerical methods in solving practically important problems of mathematical physics.

As a result of studying this course, the student should:

know: highlights of building systolic arrays and parallelization of numerical algorithms,
able: to develop practical algorithms and systolic structures parallelize the various approximate methods for solving a wide class of problems of mathematical physics.