Talk:Comparison of CPU microarchitectures

Please add date field for list

A field for architecture release date is needed (and OEM field would also be helpful.) Thanks cӨde1+6 ^LogicBomb! 18:17, 30 September 2015 (UTC)

Good idea. I added a year column. -- Frap (talk) 17:00, 6 May 2020 (UTC)

Need lede

The existing The following is a comparison of CPU microarchitectures. does not perform any of the functions given in MOS:LEAD:

It should identify the topic, establish context, explain why the topic is notable, and summarize the most important points, including any prominent controversies. The notability of the article's subject is usually established in the first few sentences. As in the body of the article itself, the emphasis given to material in the lead should roughly reflect its importance to the topic, according to reliable, published sources.

There is a link to Microarchitecture, but narrative is needed. Shmuel (Seymour J.) Metz Username:Chatul (talk) 04:21, 9 August 2020 (UTC)

need help sourcing dates, and maybe more collaboration if anyone wants to.

I am trying make this article brand-new, and in the same direction include as much available, and new, or older computer micro-architecture as possible, while summerizing, clarifying, and explaining the importance the the chips in detail. with specfifically the post as mentioned in 2020 here (https://en.wikipedia.org/wiki/Talk:Comparison_of_CPU_microarchitectures#c-Chatul-2020-08-09T04:21:00.000Z-Need_lede)

My goal end, is to make it feasible to compare everysingle architecture, and make it easy to glance at hardware features, and properly show in text, what part of a CPU, compared to others, gives it its perks, speed, technology specific features, like SIMD, superscaling specific components or what ever else can be observed to trend with processor preformance, or maybe other particularities such as specific years where a introduction in a specific FAB method directly shows some increase in chips of all vendors having a increase in performance, or something similar. That as well, Im hoping to include everysingle chip ever made, alongside everysingle time they have been used, with any particular fab method, or patented similar technology. And thanks for anyone else to help with this article! Valttang (talk) 00:39, 1 April 2026 (UTC)

I'm not sure what the scope is. Features that differ among vendors or even within a single vendor include

1. Cache
2. Hardwired versus micro-programmed
3. I-unit versus microcode instruction fetching and decoding
4. horizontal versus vertical microcode
5. Degree of parallelism within horizontal micro-instructions
6. Degree of decoding in horizontal micro-instruction fields
7. Widths of data paths and registers
8. Atomic operations in hardware
9. Channel implementation

There's a wealth of information on IBM S/360 and S/370 in CE manuals, but what bitsavers has is very incomplete. I urge anybody that still has copies to scan them. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:05, 1 April 2026 (UTC)

It looks like someone is trying to combine the instruction set architecture with the general cpu architecture ha! that sounds quite a bit difficult, but i believe we can do it, do you also have anymore information around the Elbrus series cpus, or more info regarding its sucessors as refrenced in https://en.wikipedia.org/wiki/Elbrus-8S, where the https://en.wikipedia.org/w/index.php?title=Elbrus-16S&action=edit&redlink=1 hasnt been made yet Valttang (talk) 19:42, 1 April 2026 (UTC)

also, if anyone can somehow reach out, and get frap to join us again, he can likely be a major contribution, as he was some 5 years ago, he trail marked everything so far Valttang (talk) 19:44, 1 April 2026 (UTC)

Valttang, you asked for help at my Talk page, so here are some general comments, as I am not an expert in this domain.

• As a general response, you have added two new columns, 6: Out of order execution & ROB size, and 7: Superscalar processor, without citations in any of the added cells. This is a major problem as it is non-compliant with Wikipedia's core policy of WP:Verifiability of content, and risks a rollback of all of your good-faith additions. So before you go any further, I would get busy on adding footnotes to reliable sources that back the added content.
• Note: some editors or WP:WikiProjects accept unsourced content in a specific, limited case, namely a list article where a list item is blue-linked directly to an article where the unsourced content here is directly supported there by a citation to a reliable source that "directly supports" the content here. I don't happen to agree with that approach as a proper reading of WP:V, as imho it pushes off the responsibility for verifiable content onto some other editor, which I view as a violation of WP:BURDEN. That is something you can debate here with other editors. Two things to keep in mind: it's questionable whether this is a list article (although maybe it could be seen that way, and a page move to List of CPU microarchitectures is not out of the question, as the article does not actually compare anything, afaict, it lists them, so by WP:AT the current title is a mismatch for the content) and secondly, there is no question that unlinked content must be verifiable and may be removed by anyone if it is not sourced.
• You have added triple question marks ('???') to some cells. I don't believe this is standard procedure; if something is not known, I would just leave it blank. But that is something you can discuss at a higher level than this one article; maybe at WT:SAL, or more likely at Wikipedia talk:Manual of Style/Tables. Ditto your addition of 'Maybe?'—besides the sourcing problem, how is 'Maybe' an improvement to the page, or how does it help a reader understand the information provided?

Hope this helps! Mathglot (talk) 20:55, 1 April 2026 (UTC)

I see, I frankly didnt take in much respect with using the "maybe?" in some areas, as I felt, given it was a available as a colorable tool, and its word being similar to "potentially" I simply threw away the task of verifying them for later, and used the "maybe?" as filler for it to finish later, or label genuinely unreliably answerable knowledge with it (such as the odd case of persay light based computation, would we use its focal point, or material to act using the light itself to measure its logical density size, I dont know for that yet), I do plan to trackdown or immediately detail, and show all of the sources I used for this article, and I hope to catch up, and I see i need to focus more on sourcing, so that i dont lose any progress. Beyond this article though, in your advice, would it be fair then if I move it to a list of cpu microarchitectures instead? as it, would indeed be more formal, and generally, I will attempt at adding more CPU architectures that arent really reliable to be compared, such as room computers, or micro-controllers, which arguably cant do the fraction of a sands worth of anything, and comparing them would be a waste of attention, but listing them, and using the inbuilt tools of wikipedia instead, it can be interpted by the user themselves as a comparable list for other architectures, and interesting knowledge regarding CPU manufacturs, and designers like that... meanwhile itself is simply a large list, capable of being compared with many other cpu architectures.

can you also explain the process of moving a article? i myself am quite ready to eagerly change it, as it would be a beneficial change to the article. Valttang (talk) 03:04, 2 April 2026 (UTC)

Valttang, regarding this:

I do plan to trackdown or immediately detail, and show all of the sources I used for this article...

That is a backwards approach, and far from optimal. Inline citations should be added at the same time as the content, as you should have the sources before you start writing. Normally, I would have deleted your unsourced additions, but as you are a new user, you get some slack. How long do you think you would need to source everything, would a week be enough? Please remove anything you cannot verify.

Yes, in my opinion it would be fair to move it to a list of cpu microarchitectures, but as it is possible that someone might disagree, that puts it into the category of a controversial move. A controversial move must be discussed first among editors, and only carried out after the discussion if there is clearly a consensus for it. The procedure to follow is outlined at WP:RM#CM. If you decide to start a move proposal, please start a brand new section for it. The reason the title is wrong in my opinion, is because no reliable source (or content) in the article is doing any comparing; all the comparison takes place in the head of the person reading the article. That makes the article a list, not a "comparison", per WP:Article title policy. But I am only one editor, and other editors deserve to have their opinions heard on the subject. Thanks, Mathglot (talk) 03:44, 2 April 2026 (UTC)

Thank you for your timely reply, I personally have been saving effort, with saving any potential edits in a local text file on my machine, ill go ahead and simply rollback everything to its prior original state, and begin citing sources the other editors seem to have neglected the times before, and re-add any additional information that i had added, slowly of course with citations first, before enlisting them. Instead, I will try and attempt to move the page, can I simply create a new talk page without any additional prerequisites and reach out, wait 1 month, and see if i have a good or bad consensus on it? or do i need more context and i may not, as I havent done any "moving" before. sincerely, thank you for your leniency, and giving me slack, im grateful i had this opportunity to learn from my mistake, and be allowed to change it myself, much appreciation! Valttang (talk) 03:59, 2 April 2026 (UTC)

Never mind (with specifically how to move pages), I concluded on my own, ill simply make a brandnew page, and repurpose details from this article, instead of trying to begin a exchange of decisions for what is valued for comparisons of cpu architectures, ill create a page of listing cpu architectures, so moving wont be necessary, or controversial, but ill be sure to sight sources completely with it. Valttang (talk) 04:38, 2 April 2026 (UTC)

Valttang, making a brand new page (or rather, brand new content for an existing page), otherwise known as WP:TNT, is not unheard of, but given that the entire basis of Wikipedia is collaboration, you should discuss your plans here, and get some sort of buy-in first. The work of the other 31 editors deserves respect, too. If I were you, I would lay out my plans here, and maybe {{ping}} the top editors of the page (there are only a handful) to listen to what you have to say, and provide feedback. If there is no objection to your plan, then go for it! Mathglot (talk) 05:38, 2 April 2026 (UTC)

How do I ping them?, like there user in bracket? {{}} and inside there username? and I am already taking some initiative in it, at https://en.wikipedia.org/wiki/User:Valttang/List_of_CPU_microarchitectures Valttang (talk) 07:41, 2 April 2026 (UTC)

You can ping one by one: [[User:Valttang]], [[User:Mathglot]], ... ⟶ User:Valttang, User:Mathglot, ... or all at once:{{ping|Example1|Example2|Valttang|Mathglot}} ⟶ @Example1, Example2, Valttang, and Mathglot:. Mathglot (talk) 07:56, 2 April 2026 (UTC)

thank you! yes, imma try and reach out to them. Valttang (talk) 18:04, 2 April 2026 (UTC)

I hope example1 and example2 arent online anymore LOL, they would be very confused at this sudden ping related with them. Valttang (talk) 18:06, 2 April 2026 (UTC)

• Note: follow-up commentary can be found at User talk:Valttang. Mathglot (talk) 02:24, 4 April 2026 (UTC)