Intel MKL provides functions to measure performance. This provides a way of quantifying the performance improvement resulting from using Intel MKL routines in this tutorial.
Measure Performance of dgemm
Use the dsecnd routine to return the elapsed CPU time in seconds.
Note
The quick execution of the dgemm routine makes it difficult to measure its speed, even for an operation on a large matrix. For this reason, the exercises perform the multiplication multiple times. You should set the value of the LOOP_COUNT constant so that the total execution time is about one second.
* Fortran source code is found in dgemm_with_timing.f
PRINT *, "Making the first run of matrix product using "
PRINT *, "Intel® MKL DGEMM subroutine to get stable "
PRINT *, "run time measurements"
PRINT *, ""
CALL DGEMM('N','N',M,N,K,ALPHA,A,M,B,K,BETA,C,M)
PRINT *, "Measuring performance of matrix product using "
PRINT *, "Intel® MKL DGEMM subroutine"
PRINT *, ""
S_INITIAL = DSECND()
DO R = 1, LOOP_COUNT
CALL DGEMM('N','N',M,N,K,ALPHA,A,M,B,K,BETA,C,M)
END DO
S_ELAPSED = (DSECND() - S_INITIAL) / LOOP_COUNT
PRINT *, "== Matrix multiplication using Intel® MKL DGEMM =="
PRINT 50, " == completed at ",S_ELAPSED*1000," milliseconds =="
50 FORMAT(A,F12.5,A)
PRINT *, ""
Measure Performance Without Using dgemm
In order to show the improvement resulting from using dgemm, perform the same measurement, but use a triply-nested loop to multiply the matrices.
* Fortran source code is found in matrix_multiplication.f
PRINT *, "Making the first run of matrix product using "
PRINT *, "triple nested loop to get stable run time"
PRINT *, "measurements"
PRINT *, ""
DO I = 1, M
DO J = 1, N
TEMP = 0.0
DO L = 1, K
TEMP = TEMP + A(I,L) * B(L,J)
END DO
C(I,J) = TEMP
END DO
END DO
PRINT *, "Measuring performance of matrix product using "
PRINT *, "triple nested loop"
PRINT *, ""
S_INITIAL = DSECND()
DO R = 1, LOOP_COUNT
DO I = 1, M
DO J = 1, N
TEMP = 0.0
DO L = 1, K
TEMP = TEMP + A(I,L) * B(L,J)
END DO
C(I,J) = TEMP
END DO
END DO
END DO
S_ELAPSED = (DSECND() - S_INITIAL) / LOOP_COUNT
PRINT *, "== Matrix multiplication using triple nested loop =="
PRINT 50, " == completed at ",S_ELAPSED*1000," milliseconds =="
50 FORMAT(A,F12.5,A)
PRINT *, ""
Compare the results in the first exercise using dgemm to the results of the second exercise without using dgemm.
You can find more information about measuring Intel MKL performance from the article "A simple example to measure the performance of an Intel MKL function" in the Intel Math Kernel Library Knowledge Base.
