Talk:Pentium (original)

Pentium E1000

It looks like it's canceled. [Here] And [Here] It's Going To Be The Pentium E2000, Shouldn't This Be Updated? -- From derangedfirewire --- 20:17, 9 December 2006 (UTC)

Proper MMX preciseification

Attention: I removed the expression that MMX was a "major significance". The reason is: the differences between the P54C and the P55C were indeed very little. The MMX had doubled the internal cache, and at the presentation time there were hardly any mmx-optimized applications. As a matter of fact, the transitin to MMX was very light. Initial BIOS didn't even allow the use of MMX (there were to many non-mmx Pentium in stock).--AlexProfesor 02:20, 8 Sep 2004 (UTC)

Some links to specifications (i.e. number of registers, instruction set etc.) would be nice. --HJH

Pentium vs "586" name

The "numbers can't be trademarked" issue -- how does this compare to Peugeot who registered all 3-digit numbers of the form x0y as trademarks? Is the difference that Peugeot registered then in France? Should also mention why the trademark issue arose -- Intel's 486 line had been dogged by AMD and other clone manufacturers who made similar, cheaper (better?) CPUs, and sought to register "586" as a trademark to distinguish its product. This ties in with the appearance of the ubiquitous "intel inside" sticker. -- Tarquin

I can think of many other products that have numbers as legitimate trademarks: Boeing 747, Levi's 501, Heinz 57 Sauce. There must be more to this than just simply a rejection of numerals as a trademark. Mattnelsen 15:18, 3 May 2007 (UTC)

Boeing 747, Levi's 501 and Heinz 57 Sauce are all trademarkable. 747, 501 and 57 are not. I suspect the Peugeot trademarks are all "Peugeot x0y", not "x0y". — Preceding unsigned comment added by 89.239.195.102 (talk) 07:10, 31 August 2023 (UTC)

Prior art copyright restriction: 386 > 486 > 586 is a logical sequence predictable progression by those in the field. Intel had licenced these companies as second sources for 386 and 486 cpus. So AMD could legally use 486DX4/133, after the ruling it was 5x86DX5/133. Cyrix was licensed via using TIs silicon fab as they had a blanket license from Intel. TaylorLeem (talk) 21:35, 5 July 2020 (UTC)

266 Mhz Pentium MMX?

Can anyone confirm whether there was a 266 MHz Pentium MMX meant for laptops? Crusadeonilliteracy

There definately was a pentium at 266 and 300 for mobile pc's only. here is a link to Intel's website which discusses this. http://www.intel.com/pressroom/kits/quickreffam.htm

Ahh, so my memory wasn't faulty. I saw a desktop computer for sale with a supposed 266MHz Pentium MMX in it some years ago. Crusadeonilliteracy

I suppose that the article should be modified to clarify that the 266 and 300 Mhz versions were for mobile computing only.

Why not "Intel Pentium"?

...as the title of this article? I'd rename it promptly save for any protests. --Wernher 03:11, 22 Nov 2004 (UTC)

Inconsistency

I'm not really too familiar w/ the details I'm about to address, I just know that there's an inconsistency somewhere. The article says:

P5 Pentiums used Socket 4, while P54 started out on Socket 5 before moving to Socket 7 in later revisions. All desktop Pentiums from P54C onwards used Socket 7.

I hold in my hand a Pentium 133 which just came out of a socket rather clearly marked "Socket 5." Trouble is, the table shows that the 133 speed became available with the P54C series, and hence must have used a Socket 7. So which one is wrong? Bgruber 01:31, 13 September 2006 (UTC)

This page contradicts the article, which would suggest to me that the article is wrong. When taken alongside BalusC's information, it would appear a correct phrasing would actually be:

All desktop Pentiums (excluding Pentium Overdrive) 150 MHz and faster used Socket 7.

— Aluvus ^t/c 02:17, 13 September 2006 (UTC)

I agree. I have P55C (Intel Pentium w/ MMX tech. FV80503-200 2.8V). It clearly have 320 pins instead of 321 (this is also shown on the P166 pic on this page), so it definitely can be inserted in Socket 5. I had succesfully tried it with my Socket 5 m/b and was quite surprised that, according to wikipedia, that was impossible. So I corrected the Socket 5 article and I hope this article will be corrected too.

Sorry for my english.

83.149.52.37 (talk) 05:39, 30 December 2008 (UTC)

Sorry, but installing a Socket 7 microprocessor into a Socket 5 and concluding that Socket 7 microprocessors are compatible with Socket 5 is original research and doesn't prove anything. Compatibility between sockets is more than just a mechanical fit, there are also electrical-related factors, etc. to consider. Documentation and literature from Intel should be used determine what is correct and what isn't, as they are the only reliable (or authoritative) sources. Rilak (talk) 11:42, 30 December 2008 (UTC)

Pentium model 13?

Can anybody tell me what is this?I'm using this processer on my Sony laptop.Thanks.

It's not a Pentium of the sort to which this page refers. It's a Pentium M Dothan; the CPUID of "0x6Dx" for a Dothan means that the family is 6 and the model is 13 (hex D). Guy Harris 03:23, 9 November 2006 (UTC)

Hmm.I have 1496MHz CPU shown on the System Information,but when I use the EAsy System Info(a program of the EA games used to test computers)to test my laptop,it shows only 598MHz?Why is this happening?Is it because of I'm using a Dual-core CPU?And by the way,what is the diffrance between Pentium model 13 and Pentium III? —The preceding unsigned comment was added by Vintei (talk • contribs) 17:35, 9 November 2006 (UTC).

In answer to your first question, I have no idea. In answer to your second question, the answer is "no", because Dothan isn't a dual-core CPU.

The Wikipedia is probably not the best place to use as a message board for discussing this (and I don't know what is).

In answer to your third question, see the Pentium III and Pentium M pages. Guy Harris 00:41, 10 November 2006 (UTC)

Answer to your first and second question (of which the first really is not a viable question) is that there might be speedstepping involved. Or it detects the frequency by borked cycle counting. —j.engelh (talk) 02:57, 4 December 2008 (UTC)

5th generation?

Seems like it's the 6th generation..

• 1 - 8086
• 2 - 80186
• 3 - 80286
• 4 - 80386
• 5 - 80486
• 6 - Pentium

70.231.144.169 08:02, 2 December 2006 (UTC)

I think the "5th generation" term comes from ignoring the 80186 - I don't remember whether IBM made any 80186/80188-based machines, but I don't think they did; I forget what clone makers were around at the time, but I think they went straight to the 80286 as well. Guy Harris 08:49, 2 December 2006 (UTC)

Also it seems the 186 wasn't exactly a radical change but just a replacement of some of the slower microcode with dedicated hardware. Plugwash 15:09, 2 December 2006 (UTC)

There weren't any 186-based pure PC clones due to the 186's on-board timers, DMA and interrupt controllers being mapped differently to those of the PC/XT architecture. However there were some 186 quasi-clones like the RM Nimbus and the Tandy 2000 running MS-DOS with a custom BIOS. Letdorf 22:33, 3 December 2006 (UTC).

The 80186 and 80286 belong to the same (second) generation as they were contemporary, and very similar at the implementation level. That's why Pentium would be fifth generation. HenkeB (talk) 15:06, 1 September 2008 (UTC)

Hmmm.... the 186 may have been contemporaneous with the 286 but IIRC, it was closer to the 8086 architecturally, as it lacked Protected Mode. Letdorf (talk) 15:28, 1 September 2008 (UTC).

Sure, architecturally, but not entirely regarding low level implementation techniques. Both the 80186 and the 80286 had dedicated address calculation hardware and a hardware multiplier, which made both of them significantly faster than the 8086 in many situations (even if the 80186 inherited a multiplexed bus and some other attributes from the 8086). HenkeB (talk) 02:07, 2 September 2008 (UTC)

8086, as its name don't have a 1, should be "Oth Generation", although "0th" is rarely used. 113.253.22.172 (talk) 00:30, 5 April 2014 (UTC)

There is no "0th" anything. There's the first and the second and so forth. 89.239.195.102 (talk) 07:12, 31 August 2023 (UTC)

Introduced over a range of dates?

P55C 120-150 MHz has a range of dates rather than a single date of introduction. What does this mean exactly? -- Smjg 15:11, 16 January 2007 (UTC)

Intro

The intro paragraph in the current version is

The Pentium brand name initially referred only to the fifth-generation microarchitecture of microprocessors from Intel, called Intel P5 and described here, but later it covered also subsequent generations. So, the Pentium as Intel P5, considered here, was a successor to the 486 line, and was first shipped on March 22, 1993.

Is it just me or is this an abrupt and somewhat confusing first paragraph to someone looking to simply find out what "Pentium" means?

It strikes me that were it placed a bit later (and in context) it would be fine. Here it just seems badly written.

(Also, it doesn't even mention which specific "microarchitecture's" fifth generation is being discussed. That may seem obvious to us- but it shouldn't be taken for granted). Fourohfour 14:00, 8 August 2007 (UTC)

I've reworked the intro/overview and moved some stuff elsewhere. It's not perfect- IMHO there are still some bits that arguably belong in their own sections. However, I reckon it's better organised now. Let me know what you think, thanks. Fourohfour 15:16, 8 August 2007 (UTC)

Pentium Brand

Although I agree that this article is slightly torn between covering the Pentium brand and the actual P5 Pentium line, the current version of the "Pentium brand" article doesn't do anything that isn't covered just as well in the Pentium article itself. Any thoughts? Fourohfour 15:21, 8 August 2007 (UTC)

I think that there are at least three articles mingled here: Intel 586 (for lack of a better name), Pentium brand, and the sequence of Intel implementations of x86. It would be good to make these separate articles. I created Pentium brand as a start. Much more work is needed. —The preceding unsigned comment was added by DHR (talk • contribs).

The idea behind it is good, but as it stands you're right; it needs more work, as it currently doesn't add anything. Fourohfour 20:00, 11 August 2007 (UTC)

I have moved the article to Pentium (brand) and will move some of the content too. Fourohfour 18:05, 14 August 2007 (UTC)

Latest changes to article and intro

I notice that some people disagreed with the changes I made to the article; fair enough, but I'd prefer that they'd explained any perceived problems in more depth, instead of using intermittent edit summaries.

Regarding the "redundancy, contradictions, irrelevant info; correction of casual language; addition of proper terminology, references instead of their poor repetitions, etc., etc" referred to in this edit by 141.157.253.112:-

1. I altered the original version because it did have a proper opening sentence that placed things in context. It was like coming in during the middle of an article. The change removes/reverted this opening sentence and replaces it with "The Pentium[1] brand originally referred only to the Intel's "P5" (or "Intel P5") fifth-generation microarchitecture of x86-compatible microprocessors."). Same problem; it starts discussing the Pentium brand without properly "introducing" us. As an opening paragraph, it sounds wrong. Remember that we know what a Pentium is; we should not assume that all readers do.
2. Since you attacked me for redundancy, it's ironic that "...referred only to the Intel's "P5" (or "Intel P5") fifth-generation microarchitecture..." pointlessly duplicates the "P5" name. Unless these are very official terms, there's no point in duplicating it.

In this edit set. some of the English is very bad and confusing:-

1. "(sometimes referred to also as "Pentium", because encompasing just that one brand processors)"

Huh?

1. "... Pentium M, Pentium Dual-Core (of only 2007), etc."

What is "of only 2007" supposed to refer mean? It doesn't make sense. If you want to criticise my writing, that's your choice; however, a lot of your changes were nonsensical.

(The reference refers to Intel's standardisation of names, but the relevance of this is unclear; the list only needs to include a few non-P1 Pentium names, not all of them).

I also dispute that my use of language was "casual". Please point out where you felt this was the case, because I'm struggling to see why you think that.

Anyway, I'm moving some stuff to Pentium (brand), as this article clearly suffers from trying to cover both the CPU line and the use of the brand as a whole. It makes more sense to separate them. Fourohfour 19:22, 14 August 2007 (UTC)

Improvement? (Third opinion sought)

Okay; I'm seeking a third party opinion on whether the major changes made by 141.157.253.112 ("Anon") are an improvement. (Note also Anon's changes from a previous cycle.)

I don't know if Anon will take part in this discussion. I've made a sincere attempt to address what I saw as major problems with the original opening (see above) and convert it into a proper intro/overview, but there is clearly some problem between us regarding the style of these changes.

I wouldn't mind if Anon raised his/her problems properly, but he/she simply makes vague references to my (supposed) "casual" language via edit summaries, and doesn't explain further (and doesn't contribute to the talk page).

Here are my problems with Anon; I don't think the style of the new version is an improvement. This person has criticised my writing style (which is admittedly far from perfect). Yet they themselves have made spelling mistakes, used weird grammatical constructs, included redundancy and (IMHO) generally obfuscated what is trying to be said.

He/she may have some valid concerns about my version of the article, but rather than raise them properly, Anon just makes radical changes which (IMHO) introduce more problems than they solve.

Fourohfour 20:25, 14 August 2007 (UTC)

Additional: See, this is what I have a problem with. Anon has simply removed a dab line without explanation; maybe there's a good reason for it. Or maybe he/she doesn't know or care how WP works. Who knows- why should I second-guess someone if they won't bother even doing proper edit summaries? Fourohfour 20:38, 14 August 2007 (UTC)

• I find in favour of the anon. There is a clear improvement, in many areas including content , grammer and style. It is also much easier to understand too, for non-experts like me! However I do not see many problems raised by the changes. I suggest you leave this article alone, as you seem to be too possesive of it. It is common for someone who has made a page, or made major changes to become possesive, and not want any changes. The new article is a sound description. Mattbroon 11:55, 15 August 2007 (UTC)

• I think Mattbroon was your third opinion, but if you want a fourth then I would say that as long as you keep the anon's subject matter, any grammatical changes you make should not be reverted by the anon. Bad grammar and diction can be improved, but style and content is something you all shouldn't argue over. If reverts are being made to obvious grammatical and spelling errors, it would see misguided or possibly vandalish. I don't think either is the case here. Bulldog123 17:55, 15 August 2007 (UTC)

Thank you for your contributions.

I'm not sure whether Mattbroon is referring to Anon's changes made after I posted my third party request (but before he commented!) or the previous edit set that prompted the request.

With respect, I don't see why my actions are any more "possessive" than those of Anon.

My last edits were basically a reversion... but that was because my previous edit had addressed most of Anon's apparent concerns. (In one minor area of wording I accepted he was correct and retained the change). There was no point in my making further changes in that context.

IMHO, most of my changes were for grammar reasons; I'll discuss any below in future. Fourohfour 20:09, 15 August 2007 (UTC)

References

Response from 141.157.253.112: Fourohfour, loosen up!

To see the difference compare my corrections with previous versions. How do you know that you may not be sophisticated enough to see the difference? Please, do not subject the whole world to your ego and do not proclaim yourself an ultimate judge!!! Please, do not ruin my effort only, because you feel hurt by corrections of your language. Please, accept the fact that you make mistakes like everyone else including myself.

My objective was only to make a few article headers clear for non-experts, strict, accurate, neutral, complete, and cross-referenced by removing logical errors, using consistently proper terminology, but in a simple and easily understandable way, adding missing critical details and references, e.g. about TDP, and I am practically done, though my style needs correction sometimes; my previous IP address was 151.202.72.24. —The preceding unsigned comment was added by 141.157.253.112 (talk • contribs).

You said

"To see the difference compare my corrections with previous versions."

I did.

However, the edit summary is there for a good reason; it makes keeping track of changes easier and it lets you explain the reasons for the edit.

For example, why did you remove the disambiguation line to "Pentium (brand)" at the start? Your reasoning is unclear.

To me, it looked like someone had removed a useful (and valid) disambiguation line.

"How do you know that you may not be sophisticated enough to see the difference?"

This genuinely made no sense to me, because I hadn't accused myself of lacking sophistication(!!!)

After a while I realised that you meant "How do you know that? You may not be sophisticated enough to see the difference?"

It is confusing (if not blatantly incorrect) grammar like this that made your introductory paragraph hard to read.

Accusing *me* of "lacking sophistication" is ironic; you're clearly no expert in the use of language yourself.

Next;

"Please, do not subject the whole world to your ego and do not proclaim yourself an ultimate judge!!!"

Consider this. I admitted that my edits weren't perfect. And significantly, I was the one who requested a third party opinion to resolve our dispute.

So your accusations of egotism and considering myself "an ultimate judge" are frankly laughable.

Prior to this, I had tried to engage you in discussion about our mutual problems with each others edits. No response.

Your edits were as major as mine. And you're accusing *me* of arrogance?!!!

"Please, do not ruin my effort only"

Spare me the self-pitying persecution complex. I was not trying to "ruin" your effort.

I had already changed my previous version to address your apparent concerns. I felt that the new version fixed the problems. Since I knew you were planning on re-editing, I felt that it made more sense for both of us to discuss our edits instead of wasting more time doing stuff that was going to be overwritten.

You're complaining now, but the fact is that I attempted to get your opinion and start a discussion and you were the one that ignored this.

In light of this, I simply judged your edits on their own merits. With one minor exception, I couldn't see how they were an improvement, and there was no point in attempting to rewrite my own version to address your concerns, because you wouldn't discuss them. So I reverted to my version in good faith.

What would you have done in my position?

"because you feel hurt by corrections of your language."

You repeatedly made vague references to my "casual" language, but did not explain how it was "casual".

Did you mean that my language was "imprecise"? That is a different matter altogether.

"Please, accept the fact that you make mistakes like everyone else including myself."

I accepted that you were correct in one area and acknowledged this. (Intel used the phrase "introduced on" in their text, so I accepted you changing it back to that instead of "shipped on").

"though my style needs correction sometimes"

Yes, that is what I was trying to correct, mostly. You reverted and/or made drastic changes to this. Fourohfour 19:50, 15 August 2007 (UTC)

One other thing; regarding this edit summary comment:

"Fourohfour, do not play God, your language is terrible, see my discussion posting instead, please!!"

I already explained why this "playing God" accusation is baseless.... but regarding my supposedly "terrible" language you keep banging on about... please provide an example of this.

Either way, it's a farcical accusation coming from someone who keeps making stupid grammar mistakes (see above) and can't spell "distinguished". Fourohfour 20:23, 15 August 2007 (UTC)

Points of dispute

With reference to these edits by Anon, I'd like to raise some points. If they seem longwinded, it's because I'm trying to explain my problems with the article. IMHO these are mainly issues of style and language, rather than factual disputes. Fourohfour 21:02, 15 August 2007 (UTC)

Dab line

Why was the disambiguation line removed?

It was both useful and an absolutely standard Wikipedia way of differentiating two identically-named articles covering different subjects. Pentium specifically describes the P5 chips themselves, whereas the Pentium (brand) article covers the overall use of the brand. Fourohfour 21:02, 15 August 2007 (UTC)

Strange grammar put back

"Introduced on March 22, 1993[3], the Pentium succeeded the Intel486 signifying the fourth-generation.". I tried fixing this before. It was changed. Is this saying that the Pentium signifies the fourth generation? Because that's how it can be read.

Long list?

"Later, Intel used Pentium in the names of newer generations of x86 processors branded as the Pentium Pro, Pentium II, Pentium III, Pentium 4, Pentium D. " Does the list really need to be this long? It's not complete anyway, and nor should it be; the article only needs to emphasise that the name was used for later generations, giving enough examples to make this clear. Details on subsequent use are dealt with by Pentium (branding).

Use of expression "Pentium brand" to refer specifically to original i586s

A problem with this article is that it mixes the two use of the word "Pentium"; the umbrella brand, and the original i586 "Pentium" line.

I strongly dispute that "Pentium branded CPUs" must only be taken to refer to the original line (or that the "Pentium II" CPUs don't use the Pentium brand).

Examples from a quick Google search on "Pentium brand":-

• "Fifteen years of the Pentium brand came to an end today". Clearly, "Pentium brand" isn't being used to refer to the original "P5" generation here.
• Pentium brand returns, but it's really a Core 2.
• Intel drops Pentium brand (in January 2006?!)

Specifically in the new version "...the Pentium brand refers to Intel's single core..." is (IMHO) vague. I'm not sure that a widely-used "brand" can refer to anything that specifically. It's arguable that the "Pentium II" chips include the Pentium branding. Would it not be clearer to say that the original i586 Pentiums were *named* simply 'Pentium'.

Since the potential for confusion has been shown to exist, I feel that my version was better because it made clearer that it specifically meant the "P5"/i586 line.

This shows the same problem:- "Although they shared the x86 instruction set with the P5 generation (sometimes also referred to as "Pentium"), their microarchitectures were radically different from the Pentium branded CPUs'." Fourohfour 21:02, 15 August 2007 (UTC)

Solution: Rename "Pentium (brand)" to "Pentium brands".

Fourohfour, please, try to be constructive. I do not have time for lengthy debates. In short: style is less important than content in encyclopedias, the first sentence shall introduce the topic, articles should be as simple and short as possible (use facts and references, but avoid: repetitions, adjectives, empty phrases, subjective points, etc.), facts should be listed in order of importance, short descriptions should be included for convenience (to avoid over-referencing), but long descriptions should be referenced instead, everything should be strict and logical. Do not believe everything you read, because people do mistakes, often are incompetent, because - lazy and/or lacking capacity to understand complex issues, etc., etc.

Most important, BE CONSTRUCTIVE and patient.

I make mistakes especially stylistic, but, please, do not dismiss my contribution because of it. Please, do not use my stylistic errors as a pretext to deprive readers of vital info I provide in a as strict and logical manner as I can, which may be difficult to correct, because of its density. You waist you time listing my errors in the discussion; just write silent notes in the text (<!—???-->), and I will make corrections.

—The preceding unsigned comment was added by 141.157.253.112 (talk • contribs).

This reply will be as long as I feel is necessary to answer your accusations. Important stuff is in bold if you're really in a hurry.

1. I resent being accused of not being "constructive". Aiming to improve the legibility of the opening paragraph was constructive. Attempting to engage you in discussion was constructive. Please assume good faith. Although I disagree with many of your edits, I have always assumed that you did them for what you thought were good reasons. Please extend the same courtesy to me.
2. When you say that "style is less important than content in encyclopedias", I believe you misunderstand what writing style means here. Perhaps you associate "style" with superficiality. It simply refers to a way of writing; writing style here should be concerned with legibility, understandability, organisation (i.e. accessibility) and precision. Exactly the things you were concerned about.
3. I did not try to remove your material- I tried to tidy it up. I explain my reasons for this via the summary; I wish you would too because it's a pain in the backside trying to figure out your reasons for changing text.
4. I have never "dismissed your contribution"; if you still think this, you have blinkers on. My changes to your edits were mostly issues of wording and legibility. Only when I could not see the purpose of your further edits (and you did not explain them) did I revert them.
5. I have never used your stylistic errors as an excuse to dismiss the factual content of your edits. (Unless it was so badly written that I couldn't understand what it meant).
6. It is ironic that you admit that your writing style is not perfect, and yet you appear to be lecturing me on this *and* reverting my attempted improvements to your style.
7. Discussions should be conducted on the talk page, not via article comments. At most, comments should be used to note a persistent problem that may otherwise be overlooked. In such cases, they should be short, and refer the user to the talk page if necessary.
8. Talking of comments, Why should I leave comments for you to tidy your own work up anyway? Spend more time writing it better in the first place if you think it's not good enough. You accuse me of "playing God" and being proprietorial, but what's the problem you have with me editing your comments for reasons of style/clarity?!
9. What has the title of this section "Solution: Rename "Pentium (brand)" to "Pentium brands" got to do with the rest of it?

Fourohfour 19:15, 17 August 2007 (UTC)

Rename "Pentium (brand)" to "Pentium (trademark)".

Since the Pentium trademark was not consistently used for one category of CPUs, as the Xeon or Celeron were, it is impossible to classify it as a name of homogenous and exclusive group beyond the P5 fifth-generation of CPUs, so the only possibility to classify it consistently in its broad meaning seems to be as a trademark, what it actually is.

—The preceding unsigned comment was added by 141.157.253.112 (talk • contribs).

If you are talking about the Pentium (brand) article, the name of the brand or trademark is Pentium, so it should be Pentium (brand) or Pentium (brands) or Pentium (trademark), or whatever. That's the way articles are titled on WP.

With respect, are you aware of the precise meanings and usages of words such as "brand" and "trademark", and the nuances and differences between them? I'm not, and I see no sign that you are either. So conducting a discussion and/or editing the article as if we (or you) do is silly and misleading.

I do not claim expertise in this area, but if it is a problem, perhaps we should ask others with more experience. Fourohfour 19:28, 17 August 2007 (UTC)

From 141.157.253.112: Fourohfour, all your above assumptions are FALSE!!! You had no base to instigate the discussion on the subject. Sorry, but this article actually is about nuances. I gave the structure and most of wording to intros of Wikipedia's articles about almost all Intel consumer CPUs and x86 (others fixed a few stylistic errors), and so they sound similarly. There are no others to improve on that. I am the "others" and nobody else. I offered you a privilege to decide on "Pentium brands" or "Pentium (trademark)" to be polite on one hand, and to test your capacity on the other suspecting your inability, but respecting your dedication. But your dedication is not enough, and absolutely does not justify your very inappropriate tone and attitude!!! See, how much different the article's intro became to compare with, what you wrote [sic!]. 141.157.253.112 00:33, 21 August 2007 (UTC)

Dispute resolution

As with Pentium (brand), I'm refraining from editing this article at present. This isn't an endorsement of the current version- I just don't think that making changes before the problem is solved will be constructive. Going by past experience, any changes I make- for reasons of clarity and readability- will likely be rewritten by 141.157.253.112 ("Anon").

Please don't misread this as "giving in". I still feel strongly about the issue, and after leaving this for a couple of days (to let things settle down), I'll look into more constructive ways of resolving this dispute. Fourohfour 20:41, 17 August 2007 (UTC)

Pentium 50

I am led to believe this actually existed (Intel Pentium 50Mhz Q0335 Engineering Sample, later specification number documented by Intel is Q0399). Could we not find some references for this and integrate into the article? 86.158.123.152 17:54, 31 August 2007 (UTC)

Yes there are several chips know that has a much earlier s-spec. The oldest are a Q0280 made week 28 1992, marked "A Intel-P5". If it runs at 50 MHz are not tested though. Q283,Q335 and Q336 are also found and documented. Check here http://www.chipdb.org/cat-unrated-or-mechanical-_-engineering-samples-431.htm

If these were pre-production samples devices, then they're probably only of minor notability, IMHO. Letdorf (talk) 19:18, 10 April 2009 (UTC).

Number of Transistors

I realize that process or feature size implies this, but it would be nice if the article stated the number of transistors that were on the chip. Pooua (talk) 09:23, 21 March 2008 (UTC)

Article names of Pentium and Pentium (brand)

It makes sense to me that we have one article for the family of processors, and one article for the original Pentium, but should the original Pentium be the primary subject of the article named Pentium? It strikes me that someone searching for Pentium is at least as likely to be after one of the other products (when I came here, I was most confused and it took me a while to realise why the article was referring to the Pentiums as only single core). Consider how many years the Pentium brand has been in use, and how most of that time it has referred to later models. Consider also that as time goes on, I think "Pentium" is going to most likely be identified with the range of processors as a whole, not just the original model.

Lots of trademarks refer to different products, but we don't go disambiguating them in article names with "(brand)". For example, the "Macintosh" brand has been used for a a series of computers and operating systems, but that doesn't mean we only cover the original Macintosh at Macintosh, and then have Macintosh (brand) for all the later products (or similarly for Mac OS). It would strike me as ludicrous to do things in that way. Instead, it's the original Mac which is disambiguated to Macintosh 128K.

Can we disambiguate the article for the original Pentium in some way, and so that Pentium can refer to the whole range of Pentium products? Mdwh (talk) 02:17, 29 March 2008 (UTC)

Pentium have Dual-Core now!

Pentium now have some Conroe-Based dual core cpus. Well Core 2 Duo is powerful, but pentium dual-core series is served as a minor brand, but it's much cheaper but it's still pentium! so would you guys add pentium and it's dual core stuff? —Preceding unsigned comment added by Stannhuang (talk • contribs) 01:15, 21 September 2008 (UTC)

This article only covers the first-generation Pentium and Pentium MMX processors. There is a separate Pentium Dual-Core article. Letdorf (talk) 19:14, 10 April 2009 (UTC).

Factual Accuracy

Are the clock speeds correct? They seem incredibly slow. Shouldn't they also be listed in GHz for those processors that run at that speed? —Preceding unsigned comment added by Souch3 (talk • contribs) 00:47, 27 March 2009 (UTC)

Yes, the frequencies are entirely correct. This article is about the original Pentium and Pentium MMX microprocessors that were produced from 1993 to 1999. CPU clocks are a lot faster nowadays than they were then! Have a look at List of Intel microprocessors for a historical overview of Intel CPUs. Letdorf (talk) 10:22, 27 March 2009 (UTC).

Pentium Overdrive

The article has a significant amount of content about the Pentium OverDrive. There is an entire article dedicated to it at Pentium Overdrive. Just like Pentium II does not have anything about the Pentium II Xeon, which is covered at Xeon, this article should not have anything about the Pentium Overdrive. Comments? Rilak (talk) 08:31, 4 June 2009 (UTC)

I think it's reasonable, given the close relationship, to briefly mention the Overdrive in this article, with a {{main}} link to the Pentium Overdrive article. 13:27, 4 June 2009 (UTC).

No more?

Why do I still see brand new Pc's and laptops with Pentium? This article says they don't make Pentium no more... 83.108.225.137 (talk) 13:36, 4 June 2009 (UTC)

See Pentium (brand) - this article is about the original (P5, P54, P55) Pentium processor. Later "Pentium" branded processors are quite different. Letdorf (talk) 13:41, 4 June 2009 (UTC).

I think we should rename this article to "Pentium_(P5)" to make that much clearer. Intel has started naming their current processors simply "Pentium" again, after "Pentium D" and "Pentium Dual-Core", so having this one as the main Pentium article does not make any sense. Arndbergmann (talk) 18:10, 12 August 2009 (UTC)

Moving to Intel P5 (microarchitecture)

Most of the links pointing to Pentium refer to any Pentium processor or a non-P5 processor, while the page currently is almost exclusively about the initial implmementation, which is often called the "P5 microarchitecture", "Pentium Classic" or "Original Pentium" on other pages.

In Intel P6 (microarchitecture), there is the respective information about all P6 based processors like the Pentium Pro, II, III and M, which all have separate pages as well.

I have tried to change all links to Pentium that specifically mean the original P5 to link to Intel P5 again, separating them from the others. I suggest doing this move in three steps:

• rename Pentium to Intel P5 (microarchitecture)
• adapt the first section of that article to this
• make Intel P5 a redirect to Intel P5 (microarchitecture)
• make Intel Pentium a redirect to Intel P5 (microarchitecture)
• make Pentium a redirect to Pentium (disambiguation)

Arndbergmann (talk) 16:22, 17 December 2009 (UTC)

Completely unneeded complexity, as there are no real ambiguities. Pentium, Pentium MMX, Pentium Pro, Pentium II, Pentium III, Pentium 4, Pentium M, etc, are already distinct and well defined. Its better to use terms like the Pentium family or Pentium brand, instead of hijacking the name Pentium as an abbreviation for some later processors (and intel did not talk about P3, P4 or P5 — just P6...)

Trying to impose some strict microarchitecture taxonomy on all the x86-designs tend to become a little silly; the Pentium M (called "P6 variant") differed more from the real P6, the Pentium Pro, than the Pentium differed from the 486, for instance. 83.255.35.149 (talk) 01:55, 18 December 2009 (UTC)

A major reaon that I forgot to mention is that all new products in the Pentium line are called just Pentium with no qualifier, as Intel has dropped the "Dual-Core" name this year. My guess is that most people arriving at this page are actually looking for the information that is on the "Pentium Dual-Core" page. This can be seen from some of the other sections in this talk page, as well as the visitor numbers for the two pages. I don't care if the (excellent) page gets renamed to P5 (microarchitecture), Pentium (classic) or something like that, but I feel that the page under this name here should not talk about the 15 year old product. Arndbergmann (talk) 08:06, 18 December 2009 (UTC)

I'm not so sure, WP is no catalog for current Intel offerings (although many seem to disagree...!). If users has to click twice to get to the "right" article (via a disambiguition page or otherwise), that's really a minor problem as I see it. As you probably know, it would be closer to the traditional role of an ecyclopedia to emphasise history, background, and context, rather than trying to be some kind of helper for buyers. Still, if this page gets renamed, I would suggest Pentium (original) or similar. 83.255.35.149 (talk) 09:47, 18 December 2009 (UTC)

I think we should at least rename Pentium to Intel P5 (microarchitecture) and make Pentium a redirect to Pentium (disambiguation). --Racecar56 (talk) 17:03, 22 December 2009 (UTC)

I think this move was a bad idea. If it is true that "Most of the links pointing to Pentium refer to any Pentium processor or a non-P5 processor", as Arndbergmann stated above, then what does "any Pentium processor" mean? Obviously, if people are writing articles that link to "Pentium," they must think that a "Pentium" is something that can be defined and described. However, we now have no article on Wikipedia that discusses this topic; just a disambiguation page (which is not an article) telling readers to "pick one" from a relatively long list of potential topics.

Let me give you a concrete example. Say I am reading HyperTransport and I come to the sentence "For instance, a Pentium cannot be plugged into a PCI Express bus." So I click on Pentium, and I get to the disambiguation page. What does this mean? Does it mean that none of the processors on this list can be plugged into a PCI Express bus, or only some of them, and if so, which ones? I cannot tell from this page where to find out what the generic term "Pentium" means in the context of computer processors. This is not helpful to readers. --R'n'B (call me Russ) 11:14, 9 January 2010 (UTC)

This obviously means the HyperTransport article (under the situation you describe) needs to be better written and/or rewritten to not try to use a naked generic term like "Pentium" that actually has little meaning in and of itself (as is often the case with marketing and brand names). Something along the lines of "For instance, a P5 Pentium cannot be plugged into a PCI Express bus." would be much more informative and appropriate. A similar case can be made for "Dogs are black". Obviously some types of dogs are black but certainly not all are. Uzume (talk) 00:19, 12 June 2010 (UTC)

I understand your point, but part of the reason for the move is that the old state (taking your example) was not helpful either. Like many others, HyperTransport uses Pentium as a houshold name for random processors, so linking to an article describing an over ten year old implementation of a specific processor is not helpful either. Since the article was written before Intel started reusing the name (without Dual-Core postfix) and before the release of the Pentium G6950 that actually uses PCI express on the processor, so it should be changed accordingly. In 2008, this was ok, since the article was correct both when referring to the household name and to the P5. Similar changes should probably be done to other articles arriving at the disambiguation page. For many articles, the better replacement might be to link to Pentium compatible processor.

Any articles that were linking to Pentium before and actually meant the original P5 implementation of Pentium are now linking to Intel P5, which redirects to Intel P5 (microarchitecture).

At the same time, I agree that we should do something about the Pentium page and turn that into an overview article similar to Intel Core (also Celeron and Xeon, though these don't have sub-pages), pointing to the specific processors where appropriate. There is some of that in Pentium (brand) and List of Intel Pentium microprocessors, which can be merged into a new Pentium article. Arndbergmann (talk) 03:40, 11 January 2010 (UTC)

I think Pentium (disambiguation) and Pentium (brand) probably should be merged now. List of Intel Pentium microprocessors should be kept separate as a "List of..." article. Letdorf (talk) 12:33, 11 January 2010 (UTC).

I think Pentium (disambiguation) and Pentium should be merged and Pentium (brand) should be written to adhere to the plain stand along Pentium brand--there are a number of processor models that have been released with just the plain brand including the original P5 cores. Pentium (disambiguation) can be turned into a redirect to Pentium. Since Pentium is a registered trademark, no one can define its meaning except Intel and the Pentium trademark article can encompass that. Uzume (talk) 23:11, 1 June 2010 (UTC)

Pentium (original of 1993)

This would be a resonable name for this kind of article. Other articles could be Pentium (dual core of 2009) for instance. The current "microarchitecture title" is just silly; an absurd attempt to structure concepts that cannot be structured. 83.255.38.101 (talk) 07:39, 14 January 2010 (UTC)

What about Pentium (P5), with the implicit assumption that "P5" includes P54, P55 etc? Actually I've just realised that the "(microarchitecture)" suffix in the current title implies there is another article entitled "Intel P5" and "(microarchitecture)" is required to disambiguate (per WP:PRECISION), which is not the case. Likewise for "Intel P6 (microarchitecture)". Letdorf (talk) 13:20, 24 February 2010 (UTC).

Yes, "Pentium (P5)" would be good, almost as natural as "Pentium (original)", as I see it. 83.255.42.127 (talk) 13:06, 17 March 2010 (UTC)

I think the current name "Intel P5 (microarchitecture)" is fine. "Pentium (P5)" would work as well, but "Pentium (P5 microarchitecture)" would be more explicit what P5 is about. Pcap ping 13:27, 26 April 2010 (UTC)

I have now changed the name to "Intel Pentium (P5 microarchitecture)" as a compromise acceptable to everyone, regardless of whether you want to keep the original name "Pentium" (like me), or emphasize that retroactive and non-official naming of the microarchitecture, "P5". HenkeB (talk) 20:03, 1 June 2010 (UTC)

I too think the name "P5 (microarchitecture)" is fine. There are already articles on Pentium trademark and Pentium related brands. There is precedent for the P5 naming too as it was an internal name used by Intel to refer to the microarchitecture just like P6 was used for the next microarchitecture incarnation of x86. If you look up in this talk page you can see this was already decided before. If you want to expand on one of the other Pentium related articles or clean this one up to more precisely be just the microarchitecture that would be good but do not try to make this article something it is not. There are other articles for such things (There are several lists of Intel microarchitectures that reference this among others). Uzume (talk) 23:02, 1 June 2010 (UTC)

I have now changed the name to "Original Intel Pentium (P5 microarchitecture)". Please consider the logical nature of that naming before starting an edit-war ;) HenkeB (talk) 10:39, 3 June 2010 (UTC)

Intel Pentium (P5 microarchitecture)

The trademark Pentium need not and should not be in the title of the P5 microarchitecture. Despite the fact that microprocessor cores were initially released under a brand name containing such a trademark this is not the place for that anymore than the Netburst microarchitecure should be named something akin to "Intel Pentium 4 (Netburst microarchitecture)". The Pentium trademark has been used for many different brands and those brands for many different cores from many different microarchitecures. You cause more confusion than you attempt to solve my making such changes. Please discuss on the Talk:Intel Pentium (P5 microarchitecture) page first. Uzume (talk) 21:23, 1 June 2010 (UTC)

I don't see your point. "Intel Pentium 4 (Netburst microarchitecture)" would be just fine, there are no contradictions in that naming. "Intel Pentium 4 (Willamette)" and "Intel Pentium 4 (Nortwood)" would be ok too, if these were separate articles. Moreover, the original usage of a term, name, "brand", "trademark", or whatever, deserves special emphasis. HenkeB (talk) 22:54, 1 June 2010 (UTC)

Well what about "Intel Pentium D (Netburst microarchitecture)"? (obviously the same thing as "Intel Pentium 4 (Netburst microarchitecture)" so it is not a good name)

That's what redirects are used for. HenkeB (talk) 19:54, 2 June 2010 (UTC)

"Intel Pentium 4 (Willamette)" is a different thing as Willamette was a core of Netburst microarchitecture (and there were several models with different brands applied to Willamette too).

That was my point, you could subdivide "Netburst" into separate "Willamette"/"Northwood" articles, if you want. HenkeB (talk) 19:54, 2 June 2010 (UTC)

As you can see the use of the trademark/branding only confuses the issue and does not clarify it. If you want to clean up the P5 article to move trademark and branding information out into the Pentium, Pentium (brand), or other Pentium brand articles please do so but the microarchitecture was never called Pentium until Intel lost the court case to trademark i586 and then it was only employed as a marketing branding name.

This is riculous, I dont't want to "clean up" anything, I want to restore the article to an article about the original Pentium product-family from Intel (including its internal circuitry!). Of course the "microarchitecture was never called Pentium", it wasn't called anything outside Intel, other than 586 perhaps, it did not have a formal name... The article was simply a text in the same style as the 8086, 80186, 80286, 80386, and 80486 articles. Much later, someone decided to try to coerce this into this silly over-structured and retroactive microarchitecture naming. A blatant example of narrow-minded recentism, if tou ask me ;) HenkeB (talk) 19:54, 2 June 2010 (UTC)

Your including "Pentium" in the article name only dilutes the meaning of the article and confuses the issue. If anything the name Pentium was applied later and P5 is and always was the original and most proper name of the microarchitecture. Uzume (talk) 23:25, 1 June 2010 (UTC)

Ok, I suppose you would also like to rename the 80486 article to P4 (microarchitecture), or perhaps all the x86 CPU articles? I presume the Intel 8086 would be the P0 (microarchitecture) then?... HenkeB (talk) 19:54, 2 June 2010 (UTC)

There was a near-duplicate of a past version of this article at P5 (microarchitecture), so I redirected it to this article. Whether the ultimate decision is to call this topic P5 (microarchitecture), Intel P5 (microarchitecture), Original Intel Pentium (P5 microarchitecture), or Intel Pentium (P5 microarchitecture), we only need one article on it. Spacepotato (talk) 02:06, 10 June 2010 (UTC)

The article was split because HenkeB wanted an article about the original Pentium and not about the P5 microachitecture. There is precedent on using the name P5 (Intel did use this) and all the other Intel process microachitectures use a similar naming. Also you said the P5 article was older. It was not. It had significant changes from this one. I agree some details had been added here that were not on the P5 article. Uzume (talk) 14:36, 10 June 2010 (UTC)

The P5 (microarchitecture) was different from this one mostly because it was an older version. It had a few minor changes but apart from that the articles were almost exactly the same. I don't see the point in having two articles, one on the P5 microarchitecture and the other on the original Pentium, because the two are essentially the same topic: the original Pentium used the P5 microarchitecture and nothing else did. As for what the name should be, it might be best to discuss that at Wikipedia:Requested moves. I would not oppose a move to P5 (microarchitecture) if other editors feel strongly about it. Spacepotato (talk) 19:50, 10 June 2010 (UTC)

That is not true. P5 was also used on Pentium OverDrives and are the basis for newer things like Larrabee. Something I was planning to add to the P5 microarchitecture article. Uzume (talk) 00:19, 11 June 2010 (UTC)

Well, the article already mentions the OverDrive, and I added a brief mention of Larrabee. Spacepotato (talk) 01:16, 11 June 2010 (UTC)

I don't think that Larrabee being based of P5 can be used as an argument for the split of the microarchitecture article and the P5 processor article: While Larrabee may be a derivative of the P5 microarchitecture, this is more akin to how Westmere is based on the P6. Arndbergmann (talk) 09:24, 11 June 2010 (UTC)

I never wanted the split. HenkeB suggested it on his talk page after refusing to give up putting Pentium in the name of this article. I tried it but I also agree it did not work well (but then I never thought it would really). Uzume (talk) 00:07, 12 June 2010 (UTC)

Now there is a suggestion to move one of the copies of the P5 article to Pentium (brand), which seems inappropriate. The Pentium (brand) article has always described the use of the name Pentium over all the generations and was just recently extended with more information and moved to a more appropriate place as the main Pentium article. Whatever the P5 article gets named in the end (there still seems to be a lof ot debate over that), Pentium (brand) is the wrong name. Also, we currently have two almost identical articles as P5 (microarchitecture) and Original_Intel_Pentium_(P5_microarchitecture). Please clean this mess up first, either merging the two again, or making sure that they actually cover distinct aspects and do not have duplicate content. I frankly don't see the point in a split article for these since all of the original Pentiums have the same microarchitecture and nothing else uses that. Arndbergmann (talk) 16:53, 10 June 2010 (UTC)

Well that is interesting--and strange. That said, methinks there should be a distinct Pentium brand article that is separate from a Pentium trademark/disambiguation article. It should deal with uses of the simple plain Pentium brand (like the stuff on List of Intel Pentium microprocessors) vs. other derivative Pentium trademark brands like Pentium Pro and Pentium Dual-Core, etc. Uzume (talk) 00:19, 11 June 2010 (UTC)

I undestand your desire for symmetry, but splitting the P5 article is really going too far. Over the years, Intel has moved from a model where microarchitecture, core and brand directly correspond (e.g. i386) to a complete mess where in a given microarchitecture (core and nehalem) there are many different cores (conroe, wolfdale, arrandale, ...) each of which is marketed under many names (pentium, core i3, core 2 duo, celeron, xeon, ...), so we needed to split the articles to avoid duplication. In case of P5, neither the microarchitecture nor the specific cores were ever sold under any other name, so there is no ambiguity. The technical details of every core can be covered well in the P5 article and anything related to the naming can be explained in either the P5 article or in the more general Pentium article. Arndbergmann (talk) 09:12, 11 June 2010 (UTC)

I was never really arguing for the split so much as arguing for the P5 article naming based on the fact that P5 has been used in multiple branded products (albeit most heavily in the "plain" Pentium brand naming)--namely: Pentium, Pentium with MMX Technology, Pentium Overdrive, and whatever name(s) they decide to slap on Larrabee (I agree that though it is based on P5/P54C it deserves its own microarchitecture and article). Also since other microarchitectures have and are still being sold under Pentium and Pentium derived brands, I frankly think having Pentium in the name of this article is more confusing than useful (regardless of whether the word "original" or something similar precedes it). Should this article mention Pentium--absolutely and probably heavily; but I do not think it helps to have it in the article title since that is not what the article is really about and its real topic is broader than that and "Pentium" covers this as well as other things. I agree Intel has gone from a very simplistic processor design/line model to a very large and complex one (where there is mostly a three level hierarchy: µarch/core/model and only the model gets an actual marketing brand name) and it might be considerably more appropriate to leave this as "Pentium (microarchitecure)" if Intel had not and would not continue to use "Pentium" in other--especially new--products (and the few other P5 brands also slightly helps the argument for P5 naming as well). Uzume (talk) 00:07, 12 June 2010 (UTC)

Requested move

{{movereq|P5 (microarchitecture)|Better naming: article is not about Pentium as ''per se'' but about a microarchitecture first marketed under the Pentium name after Intel failed to trademark 586; This also fits in line with other Intel microarchitecture article names as listed on these pages: [[List of Intel CPU microarchitectures]] and [[:Template:Intel processors]]}}

Resolved

 – Page moved per consensus. There was some dispute over the preferred title, but the simplest title gained most support and seems to match our naming conventions. Fences&Windows 23:18, 29 June 2010 (UTC)

Original Intel Pentium (P5 microarchitecture) → P5 (microarchitecture) — Better naming: article is not about Pentium as per se but about a microarchitecture first marketed under the Pentium name after Intel failed to trademark 586; This also fits in line with other Intel microarchitecture article names as listed on these pages: List of Intel CPU microarchitectures and Template:Intel processors Uzume (talk) 05:25, 11 June 2010 (UTC)

Survey

Feel free to state your position on the renaming proposal by beginning a new line in this section with *'''Support''' or *'''Oppose''', then sign your comment with ~~~~. Since polling is not a substitute for discussion, please explain your reasons, taking into account Wikipedia's policy on article titles.

• Support Since I made the proposal I obviously support renaming but I thought I would add the proposed naming is more precise (what the article is really about) concise (the current name is long and confuses the topic with the Pentium brands) and consistent (is in line with other Intel microarchitectural articles) than the current naming and there is precedent for using the P5 name as Intel did use this name as even included it as the first part of most of microprocessor cores based on this microarchitecture. Uzume (talk) 05:35, 11 June 2010 (UTC)

• Support I moved Pentium to Intel P5 (microarchitecture) originally, from where it was moved here. I don't think there is a perfect name for this, but P5 (microarchitecture) seems to be the best tradeoff IMHO. Arndbergmann (talk) 08:43, 11 June 2010 (UTC)

• Support: mainly for commonality with P6 (microarchitecture). I still think prefixing the title of these articles with "Intel" (e.g. "Intel P5 (microarchitecture)") is worth it to give a little more context to a bare alphanumeric designation, though. Letdorf (talk) 12:09, 22 June 2010 (UTC).

Discussion

Any additional comments:

• Not surprisingly, I suggest Original Intel Pentium, plain and simple. The current long name, as well as the article split, was invented just to keep you happy and avoid an edit-war... So, again, I see several problems with your "P5 (microarchitecture)" proposal:
  • No "P5 processor" was ever advertised to the public, what people purchased was a Pentium processor (later on specified as original, "classic", or first generation Pentium). That should be reflected in the naming of the articles. What Intel called its design during development is interesting and noteworthy, but of much less importance than what is was called in technical manuals and data sheets.
  • The "codenames" (P5, P54C, P45CQS, P54CS, P24T, P55C) stood for complete microprocessor products, not for some discrete "microarchitecture" in isolation. In traditional usage of the word, a microarchitecture excludes any high-level aspects of a processor—like an instruction set and a programming model, traditionally defined by microcode. The term therefore excludes many important details in a product (such as new x86 instructions). More importantly, that kind of strict division does not fit x86 implementations well (or risc for that matter), where the microarchitecture is heavily interviened with the programming model. Neither does it fit your suggested usage of the term very well, as I see it.
  • There were several changes in most basic x86 implementations ("microarchitectures") from product to product. Trying to impose some abstract precise "microarchitecture" on several designs is therefore largely an unneeded and contraproductive abstraction (especially so in the article headings). It's better to describe the concrete features, one by one, for the technical interested reader, along with their background and motivation. Nothing in this hinders us to mention that (for instance) the Pentium Pro and the Pentium III are closely related in the fact that they share many basic implementation strategies.
  • The fact that some other articles currently use the "Px (microarchitecture)" naming scheme is not a good argument. For instance, the P6 (microarchitecture) article is rather strange and ad hoc. It is much more natural to simply define the various x86 implementation branches via the first concrete implementations (Pentium, Pentium Pro, etc) and let later processors refer to that, along with notes on any known recent changes in strategies, technologies and circuitry.
  • The same rhetorical question once again: How come you don't want names like "P4/3/2/1/0 (microarchitecture)" for the 80486, 80386, 80286, 80186, and 8086 articles? Most of them had strange names during the development phase as well ;)
  • "there is precedent for using the P5 name as Intel did use this name as even included it as the first part of most of microprocessor cores based on this microarchitecture" What? No comprendo? What are you talking about here?

HenkeB (talk) 10:12, 11 June 2010 (UTC)

I disagree; I think it is meaningful and useful to consider the P5, P54C, P45CQS, P54CS, P24T and P55C microprocessors as being similar enough to share a common microarchitecture, which has often been referred to as "P5", for want of a better description. There was certainly a major redesign of microarchitecture between this generation of processors and the following generation, commonly referred to as P6.

However, we still have to deal with a subtle semantic difference: is this article about the microarchitecture I believe is commonly known as P5, or the series of microprocessors based on this microarchitecture? Following the WP:COMMONNAME guideline, in the former case I suggest a title like Intel P5 microarchitecture. In the latter case, I suggest Intel Pentium (P5 microarchitecture), using the usual WP disambiguation convention of appending a parenthetical description. Also including the prefix "Original" is thus redundant when so disambiguated. Letdorf (talk) 13:01, 18 June 2010 (UTC).

The difference between the microarchitecture and the microprocessor cores based on that microarchitecture is an extremely subtle one at best. The vast majority of Intel processor oriented articles are mostly split between those about technologies (i.e., a microarchitecture or a microprocessor core, etc.) and brand names (e.g., models under the Xeon, Celeron, or Core i5, etc. brands). I realize, in the past, Intel's product line was much smaller and much less complex than it is today and back then there was not much reason to split the technology from the name under which it was sold. The P5 Pentium era is when Intel started naming its microarchitecture and core projects more precisely to keep from confusing everyone, including itself (of course it did better later but most can see how confusing their 486 era was). Methinks the best article architecture for this set of topics is to have an article about P5 microarchitecture and one about the Intel Pentium brand (including references to original P5 models and more recent offerings under the plain Pentium brand). There is currently a nice Pentium article that talks about the Pentium trademark and derivative brands that serves as a great disambiguation page. There are also some articles on specific Intel microprocessor cores (e.g., Yonah (microprocessor), and Gulftown (microprocessor)). I am not sure there is enough material to warrant individual articles for each of the P5 processor cores but there are several redirects for such into sections of this P5 microarchitecture article (e.g., P5 (microprocessor), P54C (microprocessor), Tillamook (microprocessor), etc.). I do not think Intel's name needs to be prefixed to all these article names when there is no confusion (so far no one has created another P5 microarchitecture) and since the other Intel microarchitecture articles are not named with an "Intel" suffix, based on WP:AT, I propose we dispense with such to be more concise and consistent. P5 (microarchitecture) might not be the most recognizable or easy to find but naming should be a balance and it is certainly precise enough. Redirects can be used to help it be more easily found and recognizable (most of which already exist).

I do think more work needs to be done in this area. Specifically there is no plain Intel Pentium brand article (but there are ones for many of the other Pentium derived brands (e.g., Pentium Pro, Pentium M, Pentium 4, etc.). Whether this should be named Intel Pentium, or Pentium (brand) is a fine line and I am not sure I would argue against either. I do think the Pentium trademark/disambiguation article could use to be parred down a bit with material being moved from it to some of the pages it refers to.

Letdorf, thanks for joining the conversation. Once you have made your decision I encourage you (or anyone else for that matter) to cast your support or oppose vote in the Survey section above. Uzume (talk) 12:28, 21 June 2010 (UTC)

No such beast as the P50

A Pentium running at 50MHz never entered production; the Pentium 60 and 66 were the initial releases. This is a PDF of the Pentium Performance Brief from June 1997, once all of the P5 Pentiums had been released. http://www.datasheetarchive.com/indexer.php?file=DSA00372125.pdf&dir=Datasheet-021&keywords=intel+pentium+p5&database=user-highscore#. There is no Pentium running at 50MHz. Also, from personal knowledge and experience I am certain there was no Pentium 50. One may also reference Scott Mueller's Upgrading and Repairing PC's line, which has an exclusive listing of all Pentiums, there being no Pentium 50 listed. — Preceding unsigned comment added by 209.162.33.89 (talk) 19:09, 28 January 2013 (UTC)

Well, I didn't add it to the table, but added a note about it and if you check the link I added, you can see that it actually did exist - even with a CPUID. But it probably never came out on the market and therefore you can of couse discuss if it should appear here. I don't know if there are other models like this that have been designed, but never "been sold" and therefore if they should be added - or this one removed. But regarding your heading that there is "No such beast as the P50", I don't agree :-) /PatrikN (talk) 21:30, 29 January 2013 (UTC)

ES should got to listings like List of Intel Pentium microprocessors but not into a model overview, implicating a production model (unless the notes are read). --Denniss (talk) 22:28, 29 January 2013 (UTC)

So you think that it shall be removed from this page/article? Well, I agree. A little research sent me to an Intel support page about Intel Engineering Sample Processors, where I learned that the "sSpec" (Qspec) code for engineering samples start with Q and production processors usually begin with S. A lookup at CPU World's list of Intel Pentium spec numbers, shows that there is about 150 engineering samples (CPU's with a Qspec code), just for the (original) Pentium series! So, shall we remove it from this list? It's OK for me. /PatrikN (talk) 22:56, 29 January 2013 (UTC)

As you implied, we cannot be listing every single engineering sample. It would be overly verbose, not aid in greater understanding of the architecture, and would be of very niche interest to those interested in the very fine details of the processor design period. — Preceding unsigned comment added by 209.162.33.89 (talk) 17:42, 30 January 2013 (UTC)

Where is info on P5 floating point bug.

5 volt pentiums had a bug where everything after the 4th decimal was incorrect. Intel had an expensive pentium replacement program, and caused customers to buy non-Intel CPUs. As competitors you fail to mention enhanced 486 cpus eg AMD 5x86-133, IBM Blue Lighting, Cyrix and TI chips. Also note limited P5 support chips and Standards transition from 486VLB with plug and play issues (cf Cobra chipset) to PCI system standards. [PCI considers ISA bus a subset but computers like ATT Globalist the VLB (ISA bus) is master of PCI in bios.] Legal rulings in favor of Intel's competition that they were only cloning 486 which they were licensed for. Important to note Intel licensing (No P5) rather failure to license Pentium to second sources. TaylorLeem (talk) 21:24, 5 July 2020 (UTC)

The error originated from designer uploading code (microcode) for math routines but didn't verify the TaylorLeem (talk) 21:44, 5 July 2020 (UTC)

Programmable Microcode

After the 5V Pentium floating point bug intel noted adding programmable microcode so bios can install patch code before boot. I know the microcode patching is on P6 bios, and bios interogation and processor specific (intel competitors) bios settings on P5x boards. WT vs WB cache for instance can scramble data if set incorrectly. TaylorLeem (talk) 23:59, 5 July 2020 (UTC)

not all pentiums had mmx

i myself have 2 pcs with pentium that does not have mmx. they are 90 mhz and 100 mhz. they are early pentiums while mmx first arrived in the later ones. i tried to run dungeon keeper 2, which requires mmx and that said that they do not have it.84.212.107.130 (talk) 12:55, 16 October 2021 (UTC)

RAM

Please add RAM support info 99.167.109.7 (talk) 17:16, 27 January 2022 (UTC)

P5 as "the world's first superscalar microprocessor to be in mass production"

The claim that the P5 was "the world's first superscalar microprocessor to be in mass production" has been tagged as dubious.

Unless the level at which production is, in this context, deemed to be "mass production" is such that none of the earlier superscalar processors were producted in that quantity, that claim is not just dubious, it's false.

Given that the i960CA was used as an embedded processor, perhaps what that claim means is "it was the first mass-market superscalar computer used in personal computers", but, in that case, what's what it should say. Guy Harris (talk) 07:38, 10 October 2023 (UTC)

Faked Pentium chips not mentioned

During the 90s (especially in Europe, although it did also affect other regions) markets were flooded with Pentium chips that were legitimate chips internally but had been coated with some sort of hard paint/lacquer and then laser etched with an increased speed rating. The retail price difference between what seems to us today to be minor speed differences was in fact quite major - hundreds of dollars for each step up in speed grade. A large criminal organization based in Germany and Eastern Europe would buy legitimate processors and legally export them to Hong Kong or Thailand, where they would be re-coated and laser etched. The whole scheme took advantage of how the Pentium manufacturing and binning process worked.

Intel did not actually make different versions of a particular Pentium model. Chips that are on different portions of the silicon wafer just naturally have different performance characteristics for technical reasons I won't get into here. So after the dice were packaged and tested, they would run the chips at increasing frequencies until they found the upper limit that they would run stably at. Often this frequency would be in between two speed grades - let's say one was tested and remained stable until 113MHz. This chip would then be put in the "100MHz bin" and labelled and sold as a 100MHz part. However, as time went by and the particular process used was perfected, more and more chips would essentially be close to perfect. Intel still had demand for the cheaper speed grades, like the 75MHz, so often chips that were sold as 75, 90, 100, 120 etc would be stable at up to 166MHz. Enthusiasts discovered this and hence the overclocking scene took off. Criminals also took notice and so they would buy the cheaper chips, rebrand them as higher speed grades, and take in huge amounts of profit.

Eventually the police ran a sting operation (called Goldfish if my memory serves) and they raised many offices and arrested the ringleaders for money laundering, tax evasion, forgery etc. This was a big story at the time and I'm surprised it's not mentioned anywhere. I'm going to add this to my list of articles to improve, to see if I can find some newspaper articles from the time that cover the bust and details of the fakes. However, feel free to use the information I've left here to add it if I haven't gotten around to it yet. It would make a great addition to the article, I think! MrAureliusR_Talk! 09:22, 23 September 2024 (UTC)