Talk:IA-32

Qualitative vs. Quantitative Measurements

Okay, here's how it goes... "amount" is a QUALITATIVE measurement, meaning that you use the term "amount" when that of which you are speaking cannot be broken down into single numbers (like sand, you cannot have one sand, or three sands, you have a lot of or a little bit of sand). When that of which you are speaking can be spoken of in individual units, you use a QUANTITATIVE measurement (six dollars, five cats, one idiot who doesn't know basic English grammar), thus you would use the word number or other such logos. To put it plainly You do not have an AMOUNT of countable objects, you have a number.. I see this 6th grade error all the time on Wikipedia and it gets rather frustrating to realise that people can't even use their own language correctly. Why can't the English teach their children how to speak (or write, for that matter)? — Preceding unsigned comment added by 209.89.16.99 (talk • contribs) 07:11, 25 December 2007‎

TODO

• 32-bit Mode Instructions
• Instruction Explainations
• Maybe links to assembler examples?
• Links to other architectures (IA16, PPC, 68000)
• Explaination of CISC (maybe a compairison to RISC?)
• Explaination of IA32/Pentuim/MMX/SSE features (Pipelining, MultiThreading, Protected Mode, etc...)
• Maybe differences between x86 assembler language and 68000 assembler language?

Robert Lee 10:37 Nov 1, 2002 (UTC)

CISC is already explained on the Complex Instruction Set Computer page, to which CISC redirects, just as RISC is explained on the RISC page. Guy Harris 00:10, 13 January 2006 (UTC)

Pipelining is a processor feature, not a feature of the instruction set. "Multi-threading" in the sense of the OS supporting multiple threads within a single process, all running in that process's address space, isn't x86-specific, so, in the context of x86 in general or IA-32 in particular, it presumably refers to hardware threading, which is also predominantly a processor feature, exposed primarily through the same instruction set features that support multiprocessing. Protected mode isn't an IA-32 feature; it existed in the 286. A paged linear address space is, in the context of x86, a feature that first appeared in IA-32 (even if it wasn't called IA-32 back then :-)); I've mentioned it in the "Architectural features" section, but more should be written about it.

The article should probably talk about new instructions available in 32-bit mode, but not about instructions carried over from 16-bit x86. It shouldn't explain all the instructions; Intel have already published some Very Large Books doing that, and there may be a wikibook about x86 assembler. Comparing any two instruction sets in detail is probably best left to a textbook; comparing to the 68k is probably best left for a historical book. :-) Guy Harris (talk) 11:35, 28 December 2014 (UTC)

Revamping X86 and IA-32 articles

I plan on making some major changes to the contents of both the X86 and the IA-32 articles, to eliminate most of the overlap between them. I plan to make X86 the general overview of the architecture, with shorter technical descriptions, and more historical descriptives. For technical info, I will let the X86 link to the IA-32 (and also other articles) for more details. I will make IA-32 the more technical-oriented article, at least for the 32-bit side of this architecture. I will also let the X86 make links to the AMD64 article for the 64-bit technical details of this architecture.

---ykhan 05:29, 2004 May 3 (UTC)

Would it be better to merge and redirect? 84.93.243.76 09:55, 13 July 2005 (UTC)

ykhan's comments indicate why X86 and IA-32 would be separate - X86 is for a general and historical overview (see, for example, the IBM POWER and PowerPC pages), while IA-32 is for technical details of the 32-bit version of that architecture. I just added to the Apple Macintosh page, in the Hardware section, a note that some Macs now have Intel x86 processors, in the same sentence where it's mentioned that some have PowerPC processors. X86 is the right page to link to, there, as it's a similar page to the PowerPC page. Guy Harris 00:07, 13 January 2006 (UTC)

What about x86 assembly language? It appears the Department of Redundancy Department has been working overtime here.... Guy Harris 01:59, 9 January 2006 (UTC)

No —The preceding unsigned comment was added by 125.17.1.146 (talk • contribs) 22:16, 17 April 2006 (UTC).

I feel ykhan has the best approach; the articles about the platform and the language should refer to each other but be seperate and distinct. --Martyr 13:32, 14 July 2006 (UTC)

NO!!!

This is a major article, the resulting article would be too damn long if the two were combined

Post 'AMD64/EM64T/x64 merge' cleanup in this article

It seems to me, after the 'AMD64/EM64T/x64 merge' significant 64-bit information here should be relieved from this article. --Charles Gaudette 20:46, 7 September 2006 (UTC)

I just did it. Please feel free to modify my humble effort as you see fit. Basically, my POV is to treat x86-64 as a new architecture, co-equal to but derived from IA-32. I feel that this is consistent with the treatment of IA-32 as co-equal with but derived from x86. -Arch dude 03:26, 16 January 2007 (UTC)

I agree. It seems odd that half of this IA-32 article is about non-IA-32 architectures. Move all that information to some other, more relevant article -- perhaps the x86-64 article -- leaving behind a sentence or two describing the relationship between IA-32 and 64-bit architectures and linking to those articles. --68.0.124.33 (talk) 18:44, 21 January 2009 (UTC)

This page needs help

This is the IA32 article, but it spends a lot of time saying "IA64 does it [some other way]". Let's avoid IA64 entirely. There's no point.

Also from a technical view, this article's a mess; Virtual Memory doesn't just mean "so you can swap to the hard drive", and the discussion of addressable space doesn't mention ring-map limits or adequately explain PAE. —The preceding unsigned comment was added by 216.196.193.58 (talk) 23:21, 21 January 2007 (UTC).

Get rid of it

Whatever this article used to contain in 2006 or so it seems to be gone now. There's not really any useful information on this page any more. I think we should just merge what's left into the X86 article and delete the IA-32 article. — Preceding unsigned comment added by Mr. Shoeless (talk • contribs) 17:18, 10 February 2013 (UTC)

• I agree this article is mostly a WP:DICTDEF now, and as such it's better merged. But someone edit wars to bring it back... 86.121.18.250 (talk) 20:50, 6 January 2014 (UTC)
  • Hi, guys. I see this article has already established its notability, so yes, anybody has the right to re-create if he or she wishes to. In addition, such things do not go to dictionary, so no, it is not a WP:DICTDEF. So, grin and bear this article, or expand it. Whichever suits you. Best regards, Codename Lisa (talk) 01:56, 7 January 2014 (UTC)
    • We also had a similar "article"/stub on Beethoven's liver which had "established nobility" bull. 86.121.18.250 (talk) 05:27, 7 January 2014 (UTC)
      • I hate "other stuff exists" discussion. You are comparing an article on someone's liver with an article on Intel's most famous architecture? That article had no expansion potential while this one can be expanded to match that of x64. Best regards, Codename Lisa (talk) 23:16, 7 January 2014 (UTC)

P.S. x86 article has already passed 60 KB size line. (See WP:SIZERULE.) Perhaps, instead of merging, it is time import contents from that article. Best regards, Codename Lisa (talk) 23:38, 7 January 2014 (UTC)

• Hello there! In short, I'm with Codename Lisa. Calling this "dictionary stuff" is a total nonsense, as it's a very important Intel architecture, and it's still around with us as part of all current Intel and AMD CPUs. It's far away from being obsolete, despite the fact you probably can't see it. In other words, keep the article. — Dsimic (talk) 16:24, 8 January 2014 (UTC)

IA-32 vs. IA-64

I would suggest to mention the naming conflict between IA-32 (Intel x86 Architecture, 32 bit) and IA-64 (Intel Itanium Architecture, 64-bit) more clearly. Because this is just confusing, everybody starts using the abbreviations x86-32 and x86-64 to refer to the different x86 architectures and IA-64 for the Itanium architecture only. 134.28.77.173 (talk) 11:47, 20 May 2014 (UTC)

Hi. Unfortunately, those who use inventive terms like "x86-32" have already made up their mind about no caring for the status quo. They never come to read; hence we can do nothing for them. No amount of clarification will do.

Best regards,

Codename Lisa (talk) 00:36, 21 May 2014 (UTC)

Even though I don't think x86, x86-64, x64 and so forth are the architecture names, but I believe that would be a verbesserte suggestion. Ja, People might confuse that IA-64 is merely the 64-bit counterpart of IA-32, but in fact of all, they are completely different architectures, without direct relationship. IA-32 is the official name to address the fundamental architecture which x86 processors use. IA-32 is short for Instruction set Architecture 32-bit version, rather than Intel Architecture, 32-bit. The latter interpretation might be the reason why people tend to use x86 instead of IA-32, and bring out confusions. Because x86 is a metonymy usage, so it does not look too distracted when using x86-32 to refer IA-32. So I agree with you! 119.51.185.109 (talk) 11:13, 20 December 2014 (UTC)

If they are not architectures, then what are they? Chicken soup? Best regards, Codename Lisa (talk) 12:47, 20 December 2014 (UTC)

This is a good question! They are not chicken soup, they are only the metonymy terms people use to refer those kind of processors. I've already answered in my reply to the creator of this section. Though reading your words here and there, Codename Lisa, you might not be an English native speaker, but your work is excellent. If possibly, books on architecture, or Instruction Architecture, might be the best hand help you out of confusion. Thank you! 119.51.185.109 (talk) 14:39, 20 December 2014 (UTC)

Well, Ok, IA-32 and x86 surely intersect and overlap, but that's fine as IA-32 (also called i386) and x86-64 pretty much make up the x86 architecture, which itself serves as some kind of an umbrella term. IA-64 is a completely different animal, which followed the path dinosaurs took long time ago. :) — Dsimic (talk | contribs) 14:51, 20 December 2014 (UTC)

You know, Dsimic, I hope you don't judge me very badly for saying this, but I am in no hurry to accept anything that this new batch of unidentified guest editors say unless I see a very strong source. It is a tall tale, accepting that something called "Intel Architecture - 32-bit" is neither architecture nor (for the most part) 32-bit. (Especially, if the person saying it is someone who uses "Ja" and "verbesserte", mistakes metonymy with hyponymy, and then adds that "Codename Lisa, you might not be an English native speaker"! I love to be polite but not to the extent gullibility.) As much as I am entertained, I might not even dignify them with a response. Best regards, Codename Lisa (talk) 20:34, 20 December 2014 (UTC)

Hehe, sure thing I won't. :) Personally, I'm always open to changing my own points of view, beliefs and parts of knowledge (which all haven't been established randomly), but only if presented with very clear evidences (better said, reliable sources). By the way, I'm not a native English speaker, but at least I don't judge others while I keep trying to be as close to that as possible. — Dsimic (talk | contribs) 09:55, 21 December 2014 (UTC)

There is no intersect or overlap between IA-32 and x86, because they both point to the same or similar things but from different aspects. And what i386, i486, i586 and i686 point to are the versions of instruction set of x86 processors. Dsimic, so you confuse yourself with it, you might lack the experience touching with UNIX workstation. Books on computer science or getting back to college might be the best helper for you out of this confusion! Password Saeba Ryo (talk) 07:09, 21 December 2014 (UTC)

As we know, "x86" has been coined to represent 8086, 286, 386, 486, 586 (Pentium) and 686 (Pentium Pro, Pentium II) generations of CPUs and associated architectural advancements. Though, what about later processors that should've continued the numbering scheme (there's no "786"), out of which some (Pentium III and partially Pentium 4) weren't x86-64 while they extended the x86 architecture further? That's where IA-32 came in, to cover the 32-bit part of the whole "extended" x86 architecture, and to clearly distinguish x86 and x86-64 from the already dead IA-64 architecture. In other words, Intel had to swallow its pride and implement AMD's x86-64 while burying its own brand new Itanium architecture; as an architecture, x86-64 is only an extension of x86, what makes them very closely related.

Please don't get me wrong, Password Saeba Ryo, but your "you might lack the experience touching with UNIX workstation" really brought a smile on my face. :) As I've already explained above, I'll be more than happy to read one more book and correct my knowledge. Could you, please, specify which book (paper, or anything else) I should read? — Dsimic (talk | contribs) 09:55, 21 December 2014 (UTC)

Dsimic, through your words, I could see a vague picture about the meaning of that what an Architecture really is! So the books, you need to read, should be the textbooks for freshmen. You do really lack the rich experiences with UNIX workstation! Or you won't so easily to make a reply. Anyway, nice to meet you! Thank you! Password Saeba Ryo (talk) 12:01, 21 December 2014 (UTC)

Dude, I'm using (far beyond "using" as in dumb desktop usage) Linux almost exclusively since 1997 or so; in the meantime, I had additional experience with Solaris (for a while) and NetBSD (not so much). Also, I've read quite a few books on computer architecture, so I'm far away from being a "freshman" as you might imagine that. — Dsimic (talk | contribs) 12:16, 21 December 2014 (UTC)

"There is no intersect or overlap between IA-32 and x86, because they both point to the same or similar things but from different aspects." "x86" includes the 8086, 8088, 80186, 80188, and 80286, none of which implement the 32-bit version of the architecture, so "x86" includes more than just IA-32. It also includes the Opteron and so on, which implement the 64-bit version of the architecture. Guy Harris (talk) 19:36, 22 December 2014 (UTC)

If not strictly, x86 is the processor or processors, while IA-32 is the associated architecture with x86 processors, which includes not only 32-bit part but also 16-bit real and protected parts. A processor could implement one architecture, mono architecture processor; and can also implement two or more different architectures, relevant or irrelevant, hybrid architecture processor. For the latter, STI Cell Processors found in PlayStation 3 might be the good example. So from this viewpoint, IA-32 and x86 are completely different things, or the same thing at different aspects. If not forget the fact when Opteron was about releasing, AMD mentioned this kind of processors with support for the legacy software at nearly native speed, not the Architecture, AMD64. So the only confusion for x86-64 might mostly lie on the processors and the architecture(s) it realized at levels. Password Saeba Ryo (talk) 04:10, 23 December 2014 (UTC)

OK, the thing is that x86 is a general speaking; while IA-32 is a specifically speaking. Do not confuse them both.

IA-64 is another thing, the native architecture of Itanium processor, and early Itanium processors also have x86 engine to execute IA-32 applications on IA-32 mode, which is the link between IA-64 and IA-32. Later Itanium processors does not employ this IA-32 engine, so IA-32 mode is simulated by emulator within operating system. Besides, HP-UX also provides emulator for Itanium to process PA-RiSC applications. Computerfaner (talk) 12:17, 23 January 2015 (UTC)

Hatnote

This is a complementary message about last edit by] User:Nurg because space constrains made my edit summary messed up. And if User:Nurg or anybody else wants to follow up the BRD chain to it D, he or she could do the discussion here.

Basically, the edit deletes this:

{{about|the 32-bit generation of Intel microprocessor architecture|x86 architecture in general|x86}}

And inserts this:

{{Redirect|x86-32|x86 architecture in general|x86}}

First, nobody types "x86-32" or click on such a link to mean just "x86". Unfortunately, same is not correct for IA-32, thanks to a past old goof: Two years ago, there was over one million link that read "x86" and came here instead. (It is a WP:EGG violation.) So maybe deleting the old disambiguating hatnote was okay if the problem is solved. (I doubt Nurg has investigated that it was solved.) But putting the new hatnote instead is not. Second, Wikipedia policy on neologism is to not to advertise them if there is no evidence that they are used at all. I don't know who created that "x86-32" redirect but giving equal validity to it is wrong. Fleet Command (talk) 06:49, 7 December 2014 (UTC)

x86 Is Not An Architecture Name

x86 is not an architecture name, and IA-32 is not the third generation of this x86 architecture.

IA-32 is the architecture which x86 processors comply with. Even denoted as IA-32, the 16-bit real mode is part of it; for the pure 16-bit real mode, there is not official architecture name used to refer it. 221.9.19.167 (talk) 15:33, 19 December 2014 (UTC)

Hi.

Your source, please!

Best regards,

Codename Lisa (talk) 15:35, 19 December 2014 (UTC)

Generally, x86 is used to refer to the architecture which x86 processors use; and use IA-32 to refer to 32-bit part of x86 architecture, even though which also include 16-bit real and protected parts. Computerfaner (talk) 11:34, 23 January 2015 (UTC)

It would be the best to provide a reliable source, if possible. — Dsimic (talk | contribs) 11:45, 23 January 2015 (UTC)

IA-32 is the 32-bit version of the x86 instruction set architecture (ISA)? Give some strong evidences!

IA-32 is the 32-bit version of the x86 instruction set architecture (ISA)? Give some strong evidences! — Preceding unsigned comment added by 221.9.17.174 (talk) 23:01, 15 May 2017 (UTC)

Contested deletion

This page should not be speedily deleted because the reasons stated by the proposer are not among those supported for speedy deletion. Jeh (talk) 09:50, 17 May 2017 (UTC)

Merger proposal

This article has been proposed for merging with x86. If you are interested, please participate in the merger discussion. Please do not reply to this message, please discuss the merger on the other article's talk page. Thank you. Murph9000 (talk) 23:26, 17 May 2017 (UTC)

Circular reference

NB 1 in 3rd paragraph references a SuperUser thread which references back to Wikipedia. Please find a more reliable source. Thanks — Preceding unsigned comment added by 122.61.41.95 (talk) 08:15, 22 June 2019 (UTC)

The claims about "IA-64t" and "IA-32e" in the SuperUser thread don't cite Wikipedia, so that part, which is the part we're citing, isn't circular - but, then again, it doesn't cite anything else, and "somebody on a Web forum says XXX" falls under WP:SPS, so it's not a reliable source. Guy Harris (talk) 18:36, 22 June 2019 (UTC)

I completely removed the note; the bit about IA-32e being a name Intel used for x86-64 didn't seem particularly relevant to why IA-64 wasn't used, even though a reliable source exists, and I couldn't find any obvious source for the claim that IA-64t was used for x86-64. Guy Harris (talk) 18:51, 22 June 2019 (UTC)