Talk:Instruction cycle

is this even accurate? User_talk:212.219.142.161

Not clear at point 3 ("indirect address" is vague) User_talk:87.5.218.102

True, shouldn't that be direct address, anyway? Olivia Guest 14:18, 23 December 2006 (UTC)

I think this article should focus on instruction cycles in general (why they're needed, how they operate, and history) rather explain a 5 step cycle. I'm going to fork the specific cycles into separate articles, and clean up this one, if no one objects. falsedef 20:46, 16 January 2007 (UTC)

Actually, I've found other articles with the correct cycles. I'll clean this page up later. falsedef 23:21, 18 January 2007 (UTC)

It would be nice if you bring what you've already got here, as other people may start working on the same things.. sorry, just want to avoid a clash.. Coil.uk

This article is clear to me.

the fetch-execute cycle is a load of rubbish —Preceding unsigned comment added by 79.79.6.173 (talk) 22:42, 26 March 2009 (UTC)

Indeed, there is also a lot of vandalism going on, maybe someone can clean this up and perhaps lock the page? User:Anonymous 11:45, 07 March 2010 (UTC) —Preceding unsigned comment added by 91.105.203.94 (talk)

Tidy up

I'm not an expert on this topic and it's been a long time since I've studied it at school. As a means of refreshing my knowledge I've decided to read up on the topic with a view to tidying up this page. I've made a few edits already, please review for accuracy. I've tidied up a lot of grammar, I hope I haven't polluted the message. Thanks. --Eamonnca1 (talk) 19:33, 9 July 2010 (UTC)

Update - I've done a considerable amount of cleaning up here, but that diagram still doesn't tie in with the text on the page. What's the MAR and MDR? Also, I think that step 4 should be 'Interrupt cycle' rather than 'store results' based on that source that I cited later in the article. --Eamonnca1 (talk) 00:08, 10 July 2010 (UTC)

If the abbreviations MAR, MDR and CIR are to be retained and explained, the diagram needs amending follows: "Load Data Required To MDR" --> "Load Instruction to MDR" and "MAR Contents To CIR" --> "MDR Contents To CIR". AlanS1951 (talk) 11:44, 12 April 2013 (UTC)

CPUs built in the early years, such as the PDP-8, invariably have a memory address register (MAR), a memory data register (MDR), and an instruction register (IR) that you can physically touch. Because physical registers were so expensive, such early CPUs typically had a total of 5 or fewer physical registers.

Many programmers know nothing about these registers, because that knowledge is not necessary to program the computer, even in assembly language -- they are not part of the programmer-visible instruction set "registers". Those early CPUs pretend to have a dozen or so programmer-visible ISA "registers" by storing most or all of the programmer-visible registers in RAM, only loading one or two of them at a time, when needed by some instruction, into the handful of physical registers. Nowadays, CPUs still pretend to have a dozen or so programmer-visible ISA "registers", even when they have hundreds of physical registers, by using complex register renaming techniques. How can we make this article better? --DavidCary (talk) 03:45, 19 August 2011 (UTC)

Instruction cycle

Why does it start at 2?? What happened to the all important 1? Anyone? Antonyh3 (talk) 01:58, 9 November 2010 (UTC)

T-cycle

The Zilog Z80 article mentions "T-cycles". The Intel 8008 article mentions "T-states". Are they really the same thing? Should this instruction cycle explain what those states are? --DavidCary (talk) 03:51, 19 August 2011 (UTC)

According to their articles, a Z80 t-cycle is one clock cycle and a 8008 t-state is two clock cycles. So no, they're different. They are specific to those machines, and don't belong here. -R. S. Shaw (talk) 06:24, 21 August 2011 (UTC)

Original Source

I am not a regular wikipedia author and am busy writing at the moment (sorry that I am not just editing the article directly). For my research I tracked down what I believe to be source material for fetch-decode-execute paradigm. It would be nice if someone could include a citation for it:

J. Holland. A universal computer capable of executing an arbitrary number of sub-programs simultaneously. In Papers presented at the December 1-3, 1959, eastern joint IRE-AIEE-ACM computer conference, IRE-AIEE-ACM ’59 (Eastern), pages 108–113, New York, NY, USA, 1959. ACM. — Preceding unsigned comment added by Infinitesteps (talk • contribs) 20:12, 11 November 2012 (UTC)

Experts

I don't know how to call in an expert from the Wikipedia project Computer Science, but do you think this is a good idea? Halp. — Preceding unsigned comment added by 66.205.158.7 (talk) 19:50, 17 February 2013 (UTC)