Talk:Cycles per instruction

Computing Start‑class

	This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing
Start	This article has been rated as Start-class on Wikipedia's content assessment scale.
???	This article has not yet received a rating on the project's importance scale.

How to inform this is article is a stub ??? --Hgfernan 18:45, 1 September 2005 (UTC)[reply]

merge

I suggest merging Instructions Per Cycle into Cycles Per Instruction. Both talk about exactly the same thing -- after a bunch of clock cycles, how many instructions were executed?

Or does this googlefight result mean we should go the other way, merging everything into Instructions Per Cycle? --68.0.124.33 (talk) 18:28, 26 March 2008 (UTC)[reply]

They are entirely different things. CPI is for measuring the average amount of cycles it takes to complete one instruction, while IPC is used to measure the number of instructions that can be executed in one cycle. Rilak (talk) 05:01, 28 March 2008 (UTC)[reply]

Oppose, while historically C/I used to be reciprocal of I/C, since 1990s this is no longer true. Nowadays, C/I is all about increasing pipeline length to increase frequency, and I/C is all about instruction-level parallelism, multithreading, etc. They both tend to grow as technology develops. --Kubanczyk (talk) 09:19, 5 April 2008 (UTC)[reply]

Support. The current Wikipedia articles cycles per instruction and instructions per cycle both claim that one is simply the inverse of the other. If that is correct, then we should merge these two articles, the same way that miles per gallon and litres per 100 kilometres are now a single article.

Some of you seem to think that these articles, like (unfortunately) many other poorly-referenced articles, are simply incorrect, or at least outdated. I urge those of you that think this way to please correct these articles and add references that support the correct, modern usage of these terms. --DavidCary (talk) 19:08, 22 January 2011 (UTC)[reply]

Support: These are the same topic. — Omegatron (talk) 16:09, 9 December 2020 (UTC)[reply]

Are the examples relevant?

How does the fact that a jump instruction take 3 cycles to complete relevant to the time the CPU is going to spend running a program?

For example, if we have a simple infinite loop:

  label1:
     j label1

Then if we calculate like the examples seem to suggest, we get a CPI of 3 for this program (single jump instruction), when it should really be infinity:

  CPU time = # instructions * CPI / clock rate
  ==>
  infinity = 1 instruction * CPI / (some fixed clock rate)
  ==>
  CPI = infinity (not 3!)