How can we $($if we can!$)$ modify the following loops so that they can be maximally parallelized. As a first step identify and clearly describe what sort of dependencies exist in every loop $($instruction per instruction and loop iteration per loop iteration$).$ Clearly justify your answer whether the loop is parallelizable or not and if so how. Loop $1$: $(2$ pts$)$ for $($i$=0$;i$<100$;i$++)\{$ A$[$i$]=$ A$[$i$]*$ B$[$i$]$; $/*$ S$1*/$ B$[$i$]=$ A$[$i$]+$ c; $/*$ S$2*/$ A$[$i$]=$ C$[$i$]*$ c; $/*$ S$3*/$ C$[$i$]=$ D$[$i$]*$ A$[$i$]$; $/*$ S$4*/$ Loop $2$: $(2$ pts$)$ for $($i$=0$;i $<100$;i$++)\{$ A$[$i$]=$ A$[$i$]+$ B$[$i$]$; $/*$ S$1*/$ B$[$i$+1]=$ C$[$i$]+$ D$[$i$]$; $/*$ S$2*/\}$