topic-piplelined instruction processing

Upload: jassi-bhangu

Post on 05-Apr-2018

224 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    1/12

    ByAditya pal

    E1903-B56

    1Prepared By - Aditya pal

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    2/12

    P i p e l i n eIt is the set of instructions by which we can

    increase the output giving capability of anycomputer or any other digital electronicdevice.

    Pipeline increases the execution process.

    Pipeline refers to the fact that each step iscarrying data at once and each step isconnected to the next.

    Prepared By - Aditya pal2

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    3/12

    It is the simple process we were using inour computers, as we gave instructions toour computers and they were beenexecuted.

    Now in modern era instruction process issame but the feature is changed by the

    new keyword PIPELINE.

    Prepared By - Aditya pal3

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    4/12

    As the question arisesThe answer is simple in order to execute

    out commands fast and easy we had tomake our instruction process simpler andfaster.

    To make all the above easy and simplepipelining was introduced and is used nowdays by users.

    Prepared By - Aditya pal4

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    5/12

    Pipelining is a new concept and was veryeffective to the world of computer

    The origin of pipelining is thought to beeither the ILLIAC II project or the IBMStretch project. The IBM Stretch Projectproposed the terms, "Fetch, Decode, and

    Execute" that became common usage.

    Prepared By - Aditya pal5

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    6/12

    Instruction fetchFetch the instruction, retrieve operandsfrom registers Instruction decode and register fetchDecode the fetched instruction and also fetches the

    registry ExecuteExecute arithmetic instruction, compute branch targetaddress, compute load/store ,Memory address

    Memory access

    Access memory for load or store, Fetch instruction at target ofbranch

    Register write backWrite instruction results to registers

    Prepared By - Aditya pal6

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    7/12

    Prepared By - Aditya pal7

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    8/12

    Prepared By - Aditya pal8

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    9/12

    When a programmer (or compiler) writesassembly code, they make the assumptionthat each instruction is executed before

    execution of the subsequent instruction isbegun. This assumption is invalidated bypipelining. When this causes a program tobehave incorrectly, the situation is known as a

    hazard. Various techniques for resolvinghazards such as forwarding and stalling exist.

    Prepared By - Aditya pal9

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    10/12

    The cycle time of the processor is reduced,thus increasing instruction issue-rate inmost cases.

    Some combinational circuits such asadders or multipliers can be made fasterby adding more circuitry. If pipelining is

    used instead, it can save circuitry vs. amore complex combinational circuit.

    Prepared By - Aditya pal10

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    11/12

    A non-pipelined processor executes only a singleinstruction at a time. This prevents branch delays (ineffect, every branch is delayed) and problems withserial instructions being executed concurrently.Consequently the design is simpler and cheaper tomanufacture.

    The instruction latency in a non-pipelined processor isslightly lower than in a pipelined equivalent. This is dueto the fact that extra flip flops must be added to thedata path of a pipelined processor

    A non-pipelined processor will have a stable instructionbandwidth. The performance of a pipelined processoris much harder to predict and may vary more widelybetween different programs

    Prepared By - Aditya pal11

  • 8/2/2019 TOPIC-Piplelined Instruction Processing

    12/12

    Prepared By - Aditya pal12