1 votes 1 votes CO and Architecture co pipelining + – AmitPatil asked Jan 30, 2017 AmitPatil 491 views answer comment Share Follow See all 0 reply Please log in or register to add a comment.
1 votes 1 votes R2 <-- R0 + R1 R5 <-- R3 * R4 R2 <-- R5 - R2 R5 <-- R2 / R6 I/C 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 I1 IF OF ID PO WO I2 IF OF ID PO PO WO I3 IF X X X OF ID PO WO I4 IF X X X X X OF ID PO PO PO PO WO Hence Total Clock Cycle = 16 skraj answered Jan 30, 2017 skraj comment Share Follow See all 4 Comments See all 4 4 Comments reply AmitPatil commented Jan 30, 2017 reply Follow Share Since it is explicitly mentioned that no operand forwarding is used then how can we merge WO and OF.(As you did in cycle 7 and 12).? Please explain... 0 votes 0 votes skraj commented Jan 30, 2017 reply Follow Share we can reduce one stall using half - half cycle in WO and OF. bcz register R5 are store in cycle 7 and pass the value of register R5 in next instruction and instruction I3 use the value of register R5. so we can merge WO and OF in same cycle. 0 votes 0 votes AmitPatil commented Jan 30, 2017 reply Follow Share Can we always do like this when operand forwarding is not allowed??? It seems like you have used operand forwarding...I am confused... 0 votes 0 votes skraj commented Jan 30, 2017 reply Follow Share main aim of operand forwarding is to operation perform in ALU ie instruction excute(PO) Stage and result are store in ALU . next instruction are use this result in next cycle. but without using operand forwarding result are not store in ALU result are store in WB stage and reduce the one stall using half - half cycle give the result in next instruction. 0 votes 0 votes Please log in or register to add a comment.