Multi Threading In Intel Core i7 Processors | Digital Conqueror

Multi Threading In Intel Core i7 Processors

We have ended up our last article with the  Multithreading and Multiprocessing, which helps us to have a  Processor giving its best speed for execution of processes. Core i7 processors have introduce this technology which not only makes it efficient processor but also a fastest processor on planet.

But What exactly is the difference between Multithreading and Multiprocessing – The answer is here. . .

Multi-Threading

As Processor capabilities have increased, so have the demand on performance, which has increased pressure on Processor resources with maximum efficiency. This made Software Developers to scratch their heads. Instead of wasting the time for execution of a particular Task, Why not simultaneously perform some more Task? This thinking lead to the concept of threading. They began writing Operating Systems that support Running Piece of Program called Threads. Threads are small tasks that can run independently.

Multi Threading

Consider a Small Example of Fetching a particular instruction and Executing it.
This procedure consist of :

1.    Fetch Opcode (OF): Generation of Physical Address and using Address bus for accessing it.

2.    Decode Opcode (OD): The Address there is decoded to get actual Address.

3.    Fetch Instruction (FI): With the help of Address the Instruction is accessed.

4.    Execute and Write Back (EI): Instruction is Executed and the result is Written back.

Cycles ->

1

2

3

4

5

6

Instruction 1

OF

OD

FI

EI

Instruction 2

OF

OD

FI

EI

Instruction 3

OF

OD

FI

EI

Table: Use of Multi Threading

As we can see from above table that using multithreading execution of 3 Instruction has required only 6 clock cycles. But if it was not Multithreading then it might require 12 clock cycles.

So thi is how multithreading works, If you want to knwo about multiprocessing Click Here.

1 Comment

  1. Avatar
    Raj March 27, 2010

    What you have explained hers is instruction level parallelism, in other words, instruction pipelining. This can be done even in case of a single process or thread running in the system, where the instructions of the same process are pipelined.

    Reply

Leave reply

<