HP-SEE Competition

HP-SEE Computing Challenge is a competition for students interested in programming. The project allows students to take on the HP-SEE challenge and write an improved version of the code below in a variety of programming languages.

Task:

Assume that the size of the matrix variable is N × N and the code for matrix-matrix multiplication is given in the website below. Try to modify the code in a way that improves the performance.

The change can be made in C and/or assembly language, in parallel with OpenMP. The total number of floating-point operations during matrix-matrix multiplication is (2N − 1)N2. In addition to the basic (intargc, char **argv) functions, the original version of the code uses: o gettime(void) returns system time and is used for timing the code; o matfill (long N, double *mat, double val) initializes all elements of an N×N matrix mat to the value val; o matmul (long N, double *a, double *b, double *c) performs matrix-matrix multiplication of matrices a and b and stores the result in a matrix c. All matrices have the same size N × N. The change should be made to matmul (long N, double *a, double *b, double *c) to improve the code. The format and data should remain unchanged, the code should only accept one argument from the command line (the linear matrix size N), and when executed should contain information similar to:[user@ui mmm]$ ./mmm-v1 1000 1000 7.144675e+00 2.799288e+08

Code submission

An improved version of the code should be submitted to e-mail:  mmm@lists.hp-see.eu  This e-mail address is being protected from spambots. You need JavaScript enabled to view it You should also provide the following information: full name and surname, name of institution, country. One contestant can submit several versions of the code, the best of which will be taken into account. The deadline for submitting the code is October 31, 2012.

Prize

The winner of the best code will receive 900 euros. The winner must provide proof of student status to receive the prize. The winner will also be involved in various interesting HP-SEE projects and trainings. To download Multiplicationcode and for additional information, please visit the website:  www.hp-see.eu