Template talk:N-bit

2005
When browsing through the N-bit articles, it became apparent that all of them essentially used the same intro-text. However it also became apparent that the quality and completeness of these intro-texts where not consitstent.

So there were two options open:
 * 1) Create an article for "bittage" or something.
 * This was not really an option because I don't agree with thinking up new words for article names.
 * Also I think that an article is much easier to read if you don't have to follow every sublink to understand what an article is about.
 * 1) Create a parameterized template with the intro-text.
 * Generally speaking this kind of text-duplication is not really recommended.
 * However, in this case, I think it is the best way.

Even if you agree with me adjusting this template to contain intro-text, you might still not agree with the actual intro-text. Therefore I'll provide some ideas on how I could have done things differently. Maybe you think that something can be done better or neater!


 * 1) "0.5 bits" and "1 bits" looks ugly.
 * 2) "There are 28 (256) possible permutations for 8 bits." and "A 16-bit integer can store (...etc...)" could be on this page?
 * 3) See also
 * 4) *8-bit
 * 5) *16-bit application
 * 6) *16-bit era
 * 7) *32-bit
 * 8) *32-bit application
 * 9) *64-bit
 * might be redundant or should be on this page?
 * 1) Add some heading structure?

Please let me hear your ideas on this! (I'll keep this page on my watchlist.) Shinobu 09:59, 5 Apr 2005 (UTC)

Byte/octet
Thanks P3d0! Your help is much appreciated.

However, there is a minor problem:

A byte is the basic unit of data storage, wich may be 8 bits, but not necessarily so. Bytes 6 bits in size have also been in use.

An octet is a unit of 8 bits.

I think the text should reflect that. At least it should not equate bytes to octets. If I can think of a neat way to naturally incorporate that in the article, I will do so. In the meantime, feel free to so yourself if you know how. Shinobu 21:04, 5 Apr 2005 (UTC)


 * A I see what you mean. Yes, "byte" now universally means 8 bits, but it wasn't always so.  Sorry about that.  I don't have time to fix my mistake right now; please feel free to reword it yourself if you want. --Doradus 20:15, May 5, 2005 (UTC)

Not such a good idea
I really think that letting the entire intro of an article be rubber stamped like it is with this template is not necessarily a good idea. How hard is it to maintain 4 articles in a consistent manner, anyhow? I defintely think this is overusing the template concept. And bear in mind that there are various different word size machines then 8N. -- Egil 17:30, 21 Apr 2005 (UTC)
 * Apparently hard enough to allow the ##-bit articles to be unnecessarily inconsistent and incomplete. I have been thinking of other solutions, lately mainly 1) merging all in one article and 2) linking. The problem with 1) is that we are planning for more specific information on every ##-bit article. The problem with 2) is that the ##-bit articles won't "flow" as well. I personally don't tend to smile when I stumble upon an article which requires me to read other articles first.
 * The octet problem has been partially solved. If other ##-bit articles emerge, feel free to link to them in this template. Use your best judgement to decide on the layout. Cheers, Shinobu 23:56, 21 Apr 2005 (UTC)
 * The first paragraph of the template seems useful (with some editing), but the remainder feels too redundant. I'd prefer it be in a separate article.  The infobox is useful for the different word sizes, but less useful for the data sizes (there is no special name for 24-bit data objects, for example); and should that also include 6-bit and 9-bit characters, for example?  The n-bit application part is confusing.  In the MSDOS/Windows world, it's fairly clear what is being talked about; but what about the OS/360 world, where addressing went from 24 to 31 bits (or something like that -- I am not an IBM mainframe expert)? --Macrakis 12:58, 5 May 2005 (UTC)
 * I'll try to get rid of this, but care must be taken when deleting stuff because there is not a 100% overlap. Shinobu 13:52, 5 May 2005 (UTC)

Why limited to these sizes?
Why are 12-, 18-, 24-, 36-, 48-, and 60-bit architectures not mentioned in this template? There are many important historical computers with those word sizes: PDP-8 (12), PDP-1 (18), SDS 940 (24), PDP-10 (36-bit word length), CDC 6600 (60), etc. Is it supposed to be only for modern 8-bit-byte-oriented architectures? If so, it should say so somewhere. --Macrakis 12:58, 5 May 2005 (UTC)
 * You are of course correct. Maybe we do need to include them in the infobox. If so, I suggest having these 6-bit-byte systems on a seperate line. Shinobu 13:55, 5 May 2005 (UTC)

New category for bit-size related articles
I'm in the process of looking through all of the articles in the Category:Computer architecture category and trying to sort them into sub-categories because the cat has become so large. I would like to group all of the N-bit articles plus others related to bit widths such as Word (computer science), Dword, Qword, maybe Endianness, etc. into a new category that is broad enough encompass such concepts.

I need help finding a proper name for the category and am asking for suggestions. Some names I have thought of (I like some better than others!) are:
 * Category:Data unit
 * Category:Word size
 * Category:Bit width
 * Category:Data bits

This new category would be placed in these categories:
 * Category:Units of information
 * Category:Computer memory
 * Category:Digital registers
 * Category:Computer architecture (? – perhaps not; the above two are already subs of architecture)

Suggestions would be helpful! JonHarder 23:07, 16 July 2006 (UTC)
 * For the record, lacking any discussion, I have created Category:Data unit. JonHarder 14:46, 28 July 2006 (UTC)

"Processor"
Shouldn't the term "processor" used in this template be changed to the less ambiguous "microprocessor" or even "CPU"? All the articles that include this template refer to computer architecture, not DSPs or SoCs. -- uberpenguin


 * I agree. Go for it.  --Doradus 12:13, 15 August 2006 (UTC)

'Word' is not same as '16-bit'
The following comment was moved from Talk:Word (computing):


 * I noticed that the 64-bit architecture page has a side panel which allows you to click on 16-bit data size link, which links directly to this page. I think that's anathema to the idea that a word-size is defined by the architecture.  There should be a page that redirects from 16-bit data size to Word (computer science).  This way, the link from 64-bit architecture does not lend the web surfer the wrong impression that 16-bit data size is always equal to word-size -- anonymous jan 18 2007 2:58 pm est —The preceding unsigned comment was added by 24.186.76.45 (talk) 19:56, 18 January 2007 (UTC).

I agree the implied connection between 16-bit and word that the template seems to make is misleading and/or confusing. (The two are often considered connected for x86-derivatives, but not for computers generally.) While updating the table for other reasons, I've delinked the 16-bit data size entry so that this doesn't happen. -R. S. Shaw 05:22, 19 January 2007 (UTC)

Address and data buses?
Just because a CPU happens to have an N-bit address bus or an N-bit data bus, that doesn't mean it's an "N-bit CPU" in as meaningful a sense as the sense in which a CPU with N-bit virtual addresses and N-bit registers is an "N-bit CPU". See, for example, Talk:64-bit. Guy Harris (talk) 22:46, 19 November 2007 (UTC)

Unexplained deletions
I've reverted two recent edits which have deleted the "boilerplate" intro text and various parts of the table as I can't see a good reason for doing this - if there is one, please discuss it here first. Letdorf (talk) 15:18, 24 December 2008 (UTC).

Longlasting problems from ancient template
This template is bogus in the way combines a stilted lead with the nav/info box. They need to be unbundled, so we can work on more sensible leads (and titles) for these articles. See Talk:64-bit. Dicklyon (talk) 22:27, 31 July 2012 (UTC)

Ah, I see it transcludes Template:Computer architecture bit widths. So we can just use that instead where we want to replace the lead. Dicklyon (talk) 03:26, 1 August 2012 (UTC)


 * Also, the bit about generations of computers is a bit bogus. It used to be included for all power-of-2 sizes, but there was never an era where 2^0-bit processors were "the norm"; those processors are very specialized.  (Then again, that page isn't 1-bit, it's 1-bit architecture, so the template's first sentence is also wrong for that page; I've extracted the relevant bits, edited them to reflect that the instructions and program counter of real-world "1-bit" machines weren't 1 bit wide, and replaced the use of  in 1-bit architecture with them.)


 * Perhaps that sentence should be completely removed from the template, and inserted, with the proper editing, into those pages where it actually applies. ("With the proper editing", for example, would take into account that during the "8-bit era" of microcomputers, minicomputers were mostly 16-bit with some 18-bit and 24-bit machines, and mainframe computers were mostly 32-bit, with some 36-bit and 48-bit machines, and that during the "16-bit era" of microcomputers, 32-bit superminicomputers were starting to appear.) Guy Harris (talk) 08:53, 1 August 2012 (UTC)


 * OK, I yanked that sentence, and put a modified version into the 8-bit, 16-bit and 32-bit pages, noting that it only applies to microcomputers and the microprocessors used in them. (64-bit computing has something closer to the original, but, by that time, just about everything, including IBM System z mainframes, was made out of microprocessors, so it's not as if non-microprocessor-based machines were ahead of microprocessor-based machines in bit width, and no "micro-" qualification is needed.  Then again, 64-bit computing doesn't use the template for a variety of reasons, including the fact that the article's title is now 64-bit computing, not 64-bit.) Guy Harris (talk) 09:30, 1 August 2012 (UTC)