A theoretical reflection...

IsMannen

New Member
... by a newbie I should add. :P I have no more knowledge in computers than an average youth.

Hello!

I will try to express my question as simply as possible.


As far as I can understand, one divides the operation of a CPU in four steps. The first one is Fetch, the second one decode, the third execute and the fourth writeback. That is correct, right?

Could one then consider the operatione of a CPU as two major parts. One part where the information in some way is processed in the CPU, and the other part when the information is sent in or out of the CPU?

My question is: Is there some way to determine (at least an aproximate value) of how big ammout of information is sent in and out of the CPU per time unit. In kb/s or kB/s? Is this value perhaps proportional to the clock frequency?..


I hope I am not TOO lost in my thinking over here...

I would be very very grateful for an explanation in either case.

Best wishes from a cold Sweden
/Oscar
 
... by a newbie I should add. :P I have no more knowledge in computers than an average youth.

Hello!

I will try to express my question as simply as possible.


As far as I can understand, one divides the operation of a CPU in four steps. The first one is Fetch, the second one decode, the third execute and the fourth writeback. That is correct, right?

Could one then consider the operatione of a CPU as two major parts. One part where the information in some way is processed in the CPU, and the other part when the information is sent in or out of the CPU?

My question is: Is there some way to determine (at least an aproximate value) of how big ammout of information is sent in and out of the CPU per time unit. In kb/s or kB/s? Is this value perhaps proportional to the clock frequency?..


I hope I am not TOO lost in my thinking over here...

I would be very very grateful for an explanation in either case.

Best wishes from a cold Sweden
/Oscar

Yes, it's called IPC (instructions per cycle)
http://en.wikipedia.org/wiki/Instructions_per_cycle

When you double click on an icon to run a program or whatever, here is what happens:

1. The program, which is stored inside the hard drive, is transferred to the RAM. The program is made up of a series of instructions to the CPU.
2. The CPU, using a circuit called memory controller, loads the program data from the RAM memory.
3. The data, now inside the CPU, is processed.
4. What happens next will depend on the program. The CPU could continue to load and executing the program or could do something with the processed data, like displaying something on the screen.

CPUs use cache to decrease the time taken to access the RAM.
http://en.wikipedia.org/wiki/CPU_cache

I may be wrong, but that's what I always thought.
 
You kids will be the death of me! You cannot compare clock frequencies across brands or generations of CPU's. Yes you can bench the output of the CPU. Clocks only matter when comparing like CPU's.
 
IsMannen;1585156My question is: Is there some way to determine (at least an aproximate value) of [B said:
how big ammout of information is sent in and out of the CPU per time unit. In kb/s or kB/s[/B]? Is this value perhaps proportional to the clock frequency?..

This was the question.

You kids will be the death of me! You cannot compare clock frequencies across brands or generations of CPU's. Yes you can bench the output of the CPU. Clocks only matter when comparing like CPU's.

Who was comparing clock cycles of different CPUs to cause your death? Or was that two statements within itself to answer a question that was never asked?
 
Last edited:
StrangleHold was the user name Pedantic already taken? I can tell I am going to like you. I am having a little trouble answering the obvious on this one.

But let me have a go.......
Is this value perhaps proportional to the clock frequency?..
Can you give your enlightened response as to how I am misreading that?:P:D
 
... by a newbie I should add. :P I have no more knowledge in computers than an average youth.

Hello!

I will try to express my question as simply as possible.


As far as I can understand, one divides the operation of a CPU in four steps. The first one is Fetch, the second one decode, the third execute and the fourth writeback. That is correct, right?

Could one then consider the operatione of a CPU as two major parts. One part where the information in some way is processed in the CPU, and the other part when the information is sent in or out of the CPU?


My question is: Is there some way to determine (at least an aproximate value) of how big ammout of information is sent in and out of the CPU per time unit. In kb/s or kB/s? Is this value perhaps proportional to the clock frequency?..


I hope I am not TOO lost in my thinking over here...

I would be very very grateful for an explanation in either case.

Best wishes from a cold Sweden
/Oscar

The first post does seem to speak about processors in general, there isn`t a reference comparing say intel and amd, or even one dual core against another.
 
... My question is: Is there some way to determine (at least an aproximate value) of how big ammout of information is sent in and out of the CPU per time unit. In kb/s or kB/s? Is this value perhaps proportional to the clock frequency?..

But let me have a go.......
Can you give your enlightened response as to how I am misreading that?:P:D

The question is to determine data passed in a clock cycle, and was it proportional to clock speed as a general statement as far as any processor goes.

Pretty much a no brainer that a Intel Sandybridge will be faster clock for clock then a AMD K6 or any two different CPU architectures.
 
Back
Top