Presented by: Sergio Ospina Qing Gao
Dec 31, 2015
ContentsContents
♦ 12.1 Processor Organization♦ 12.2 Register Organization♦ 12.3 Instruction Cycle♦ 12.4 Instruction Pipelining
12.1 Processor Organization
♦ Fetch Instruction♦ Interpret Instruction♦ Fetch Data♦ Process Data♦ Write Data
Remember….
♦ The processor needs to store some data temporally.
♦ It must remember the location of the last instruction so that it can know where to get the next instruction.
♦ It needs to store instructions and data temporally while an instruction is being executed.
Control & Status Registers
♦ Program Counter
Contains the address of an instruction to be fetched.
♦ Instruction Decoding Register
Contains the instruction most recently fetched.
♦ Memory Address Register
Contains the address of a location in memory.
♦ Memory Buffer Register
Contains a word of data to be written to memory or the word most recently read.
12.3 Instruction Cycle
♦ It is the time in which a single instruction is fetched from memory, decoded, and executed.
♦ An Instruction Cycle requires the following sub cycle:
Instruction Cycle♦ Fetch Reads the next instruction from memory into the
processor.
♦ Indirect Cycle (Decode Cycle) May require memory access to fetch operands,
therefore more memory accesses.♦ Interrupt
Save current instruction and service the interrupt.
♦ Execute
Interpret the opcode and perform the indicated operation.
Data Flow (Interrupt Diagram)
ControlUnit
PC
MBR
PC
ControlUnit
MARMAR
Memory
MBR
MemoryMemory
ControlUnit
ControlUnit
PC
Data Flow (Execute)♦ May take many forms♦ Depends on instruction being executed♦ May include
Memory read/writeInput/OutputRegister transfersALU operations
12.4 Instruction Pipelining
♦ Instruction processing is subdivided: Fetch/ Execute instruction
♦ Pipeline has two independent stages: 1st Stage – Fetch an instruction and buffers it.
2nd Stage – Temporarily free until first stage passes it the buffered instruction.
While the second stage is executing the instruction, the first stage fetches and buffers the next instruction.
♦ Instruction prefetch or fetch overlap. - Purpose? To speed up instruction execution.
Instruction Processing♦ Fetch instruction (FI)♦ Decode instruction (DI)♦ Calculate operands (CO)♦ Fetch operands (FO)♦ Execute instruction (EI)♦ Write operand (WO)
♦ Successive instructions in a program sequence will overlap in execution.
Six-Stage CPU Instruction PipelineThe logic needed for pipelining to account for branches, interrupts, and arising problems.
Branches
♦ Branch- group of instructions
♦ Branch Instructions – (Jump Instruction) One of it’s operands is the address of the next instruction to be executed.
Branches
♦ Two Types of Branch Instructions• Unconditional – Branch always happens• Conditional – Branch only happens if
certain condition is met.- The PC is updated to the address specified in the operand of the conditional branch instruction.
- A conditional branch instruction is similar to an if statement.
Conditional Branch Instructions
♦ Condition CodesBRP X
○ Branch to location X if result is positiveBRZ X
○ Branch to location X if result is zeroBRE R1,R2,X
○ Branch to location X if contents of R1 = R2
Dealing with Branches
♦ A major problem in designing an instruction pipeline is assuring a steady flow of instructions to the initial stages of the pipeline.
♦ Since conditional branches alter the steady flow of instructions, we must come up with ways to execute them efficiently.
Dealing with Branches
♦ 5 Approaches to Dealing with Conditional BranchesMultiple StreamsDelayed BranchPrefetch Branch targetLoop BufferBranch Prediction
Dealing with Branches
♦ Multiple Streams (IBM 370/168 and IBM 3033)○ Pipeline fetches both instructions.
Leads to contention delays, and branches can lead to too many streams.
♦ Delayed Brancho Branch Instruction occurs later than desired.
♦ Prefetch Branch Target (360/91 IBM)o The target of the branch is prefetched, along with the
instruction following the branch, so if the branch is taken this will speed up performance.
Dealing with Branches
♦ Loop buffer ( Motorola 68010)Memory containing the n most recently
fetched instructions.Useful with if-then and if-then-else
statements, as well as loops♦ Branch Prediction
Different techniques are used to predict whether the branch will be taken or not
If the prediction is correct this will speed up performance
Review Questions1. What are the major components of a processor? Arithmetic and Logic Unit (ALU) and the Control Unit (CU).
2. What is the function of the ALU? The ALU does the actual computation or processing of data.
3. What is the function of the control unit? The control unit controls the movement of data and instructions into and out of
the processor and controls the operations of the ALU.
4. What are the two roles that registers in the processor perform?
User-visible registers, and control and status registers.
5. What are bits set by the processor hardware as a result of operations?
Condition codes.
Review Questions (Continued)
6. What is an instruction cycle? It is the time in which a single instruction is fetched from memory, decoded, and
executed.
7. What are the four sub cycle of an instruction cycle? Fetch, Indirect (if any) , execute, and interrupt (if any).
8. Is the fetch or execute cycle the same for all CPU? No, it depends on the CPU’s design.
9. What is the sequence of an interrupt cycle? PC MBR
Address of Stack MAR
MAR Memory
PC Memory
Control Unit request memory write via Control Bus
PC is loaded with address of Interrupt handler.
Review Questions (Continued)
10. What is the main purpose for instruction pipelining? To speed up the instruction execution rate.
11. How can you make the pipelining more efficient? To gain further speedup, the pipeline must have more stages for
decomposition.
12. What is a condition code? A statement that if true will allow the branch to be executed.
13. What is another name for a branch instruction? A jump instruction.