The following Fortran program is

to be executed on a uniprocessor, and a parallel

version is to be executed on a shared$-$memory

multiprocessor.

$L\mathrm{.}1$: $Do10I=1,1024$

$L2$: ${\displaystyle \underset{?}{\overset{?}{}}}(I)=0$

$L\mathrm{.}3$: $Do20J=1,1$

$L4$: $20,{\displaystyle \underset{?}{\overset{?}{}}}(I)=$sun$(I)+I$

$L5$: $10$ Continue

Suppose statements $2$ and $4$ each take two

machine cycle times, including all CPU and memory$-$

access activities. Ignore the overhead caused by the

software loop control $($statements L$1,$ L$3,$ and L$5)$

and all other system overhead and resource con$-$

flicts.

$($a$)$ What is the toal execution time of the

program on a uniprocessor?

$($b$)$ Divide the outer loop iterations among

$32$ processors with prescheduling as follows:

Processor $1$ executes the first $32$ iterations

to $32).$ processor $2$ executes the

next $32$ iterations to $64$