PROGRAM MAT_MUL
C
C      On an SGI Origin 2000 compile with:
C      (-apo = automatic parallelization option)
C
C      f90 -r8 -apo -r10000 -O3 -col120 -o mat_mul.exe mat_mul.f
C
       INTEGER N, GETENV, OMP_NUM_THREADS
       EXTERNAL GETENV
       CHARACTER ENV_VALUE*8
C
C
       READ (5, *) N		! Array NxN
C
       ISTAT = GETENV('OMP_NUM_THREADS', ENV_VALUE) ! Get value of OMP_NUM_THREADS Env_var
       READ(ENV_VALUE, FMT='(I8)') OMP_NUM_THREADS  ! Converts int strg to int binary
C
       CALL ARRAY_MULT(N, OMP_NUM_THREADS)
C
       STOP
       END
C ---------------------------------------------------------------------
C
       SUBROUTINE ARRAY_MULT(N, N_CPUS)
       INTEGER N, N_CPUS, NSQ
       REAL (KIND=8)  SECS, TM1, TM2, TIMEF
       REAL (KIND=8)  AA(N,N), BB(N,N), CC(N,N), PI
       EXTERNAL TIMEF
C
       PI = 4.0D1 * ATAN(1.0D1)
       NSQ =N*N
       ANG_INCR = (2.0D1 * PI)/NSQ
C 
       DO J=1, N
          DO I=1, N                     ! Compute AA & BB
             AA(I,J) = COS( FLOAT(I+(J-1)*NSZ)*ANG_INCR )
             BB(I,J) = SIN( FLOAT(I+(J-1)*NSZ)*ANG_INCR )
          END DO
       END DO
C
       TM1 = TIMEF()	! elapsed real_time (millisecs): Start counting
C
       DO J=1,N
          DO K=1, N
             DO I=1, N
                CC(I,J) = CC(I,J) + AA(I,K)*BB(K,J)
             END DO
          END DO
       END DO
C
       TM2 = TIMEF()
       SECS = (TM2 - TM1) * 1.E-3
C
       WRITE(6,*)
       WRITE(6,*) 'NUMBER OF CPUS/TASKS ASSIGNED = ',N_CPUS
       WRITE(6,*) 'Parallelized Matrix Multiplication for ',N,'x',N,' Size'
       WRITE(6,*)
       WRITE(6,*) 'Elapsed Wall Clock Time ',SECS,' Seconds'
       WRITE(6,*)
C
       PRINT *, CC(1234,1), CC(8023,1), CC(2345,1)
C
       RETURN
       END