Template:Did you know nominations/Intel 5-level paging

The following is an archived discussion of the DYK nomination of the article below. Please do not modify this page. Subsequent comments should be made on the appropriate discussion page (such as this nomination's talk page, the article's talk page or Wikipedia talk:Did you know), unless there is consensus to re-open the discussion at this page. No further edits should be made to this page.

The result was: rejected by BlueMoonset (talk) 03:16, 19 July 2018 (UTC)

Intel 5-level paging edit

(

)

A diagram of five levels of paging

... that Intel plans to make its processors walk further? Intel
- ALT1:... that Intel plans to make its processors support 128 petabytes of memory using five page tables (illustrated)? ZDNet
- ALT2: ... that Intel plans to make its processors walk further (fields pictured)?

Created by Bellezzasolo (talk). Self-nominated at 22:02, 26 April 2018 (UTC).

Article is long enough and has been moved from the Draft namespace on the indicated day. Bellezzasolo has 2 DYK credits so far, no need for a QPQ. I'll still have to check for copyvios. For now I see the following issues:
1. The article is currently an orphan. I guess this needs to be resolved before it can appear on the main page.
2. I like the original hook. It is quirky but I'm not sure it is correct: As far as I can tell, it is not a processor walk but a page table walk. I'm not knowledgeable about this topic, and I haven't got access to the ACM DL source, but it seems to me (from an unreferenced section of Page table) that this process is not directly, or not always, handled by the processor. Could the article section be expanded to clarify this?

--Pgallert (talk) 16:29, 29 April 2018 (UTC)

@Pgallert: I've fixed the orphan issue, it was a natural addition to the IA32E page. Page table walks are performed by the MMU (Memory Management Unit), which is generally a part of the processor. (I mean, processors do a load of different things all at the same time). I'll clarify on the article. (Edit: I think done). ∰Bellezzasolo✡ Discuss 12:49, 30 April 2018 (UTC)

@Bellezzasolo: Ja, and someone undid your edit. I have re-inserted it in another form now. Doing the plagiarism check tomorrow morning, too much to download from my current connection. --Pgallert (talk) 17:32, 30 April 2018 (UTC)

I've added ALT2, which is the hook, but with a quirky tie in to the picture. ∰Bellezzasolo✡ Discuss 01:32, 1 May 2018 (UTC)

@Bellezzasolo: I have now done the copyvio check and detected no problems. All meticulously reworded. In fact, I couldn't find one statement in the source, details on the talk page. ALT2 is a good suggestion, but the word "fields" would have to appear in the article for this hook to be good. The picture is self-created by the nominator and can be used for the hook, ~~although without the picture subscript.~~ I have missed that this is now common. Pgallert (talk) 19:48, 1 May 2018 (UTC) Taking quirkiness a step further, how about ALT3 below? --Pgallert (talk) 09:47, 1 May 2018 (UTC)

ALT3: ... that Intel plans to make its processors walk further into the fields (fields pictured)?

Comment Fails criteria - there is no way it is of interest to a broad audience, not by any reasonable definition of that term. Only specialists in Intel CPU architecture will have the slightest clue what this is about. "Will make Intel processors walk farther" is a cute hook but that's not enough; you don't want a DYK to leave readers wondering "what was that about? I never heard of any of that stuff." Anyway, although this particular article is new, Intel has made their CPUs "walk farther" twice before (once with PAE, once with em64t paging). If it's ever actually implemented this will be completely invisible to the average user and even to most developers - hardly a broad audience. And regarding that point, the suggested hook is misleading, even just plain wrong; the reference is to a specification for a possible future implementation. There is nothing anywhere to suggest that Intel is actually going to implement it. So at this time it is a gross overstatement to say that Intel is "planning" this. Correcting the hook will only narrow the interest further. Jeh (talk) 00:05, 2 May 2018 (UTC)

This nomination has been stalled a bit by side discussions, here, on the nomination talk page, and on the article talk page. Trying to sum up, and from my perspective, we have the following:

Most typical formal DYK requirements have been checked by me and are fulfilled. No copyvio, image suitable, long enough, new enough, referenced, etc.
There is a concern by Jeh that the hook statement is not interesting to a broad audience and that it thus violates 3a of the eligibility criteria. I'm not sure how to handle this, or if the various songs, fungi, sports results and low-level biographies fare any better.
There is a further concern about the word "planned" in the various hooks. I'd like to solicit opinions and suggestions about how to make a more realistic statement, preferably keeping a somewhat quirky hook.

I'll weigh into the discussion but am currently uncomfortable to approve or disapprove the nomination in general, or a specific hook. --Pgallert (talk) 19:31, 2 May 2018 (UTC)

Regarding planned, the fact that intel developers added the code required to handle this to the Linux kernel is very indicative. The impression given by the secondary sources is that this change is in the pipeline, so I don't see anything wrong with planned. Planned extensions may not come to pass, and still be planned - and notable (e.g. SSE5). I didn't say "upcoming" because of this uncertainty, and "proposed" may be accurate, and would work in the hook. However, "planned" does have a basis in our primary source - "This document describes planned extensions". From WP:CRYSTAL, "Wikipedia does not predict the future. All articles about anticipated events must be verifiable, and the subject matter must be of sufficiently wide interest that it would merit an article if the event had already occurred" (emphasis added). This would merit an article if it comes to pass or it doesn't, so shouldn't be a problem on that front. As for interest, support for lots more RAM is an important advance, and will have be significant in the future, even if very much behind the scenes. ∰Bellezzasolo✡ Discuss 21:45, 2 May 2018 (UTC)

Regarding interest, 3a applies to the hook. We don't want hooks to be dry and boring, but to rather entice people in. Obviously subjective, and I'll leave that to others. But it only applies to the hook. The purpose of the articles, stated as an aim of DYK, is

To present facts about a range of topics which may not necessarily otherwise receive Main Page exposure
To highlight the variety of information on Wikipedia, thereby providing an insight into the range of material that Wikipedia covers.

I'd say that this means we cover somewhat quirky, niche topics, not shun them. ∰Bellezzasolo✡ Discuss 21:53, 2 May 2018 (UTC)

I don't think we should say it's "planned" unless and until it appears on an Intel CPU roadmap.

"Support for lots more RAM is an important advance." But this isn't about RAM! 4-level paging already supports a 52-bit physical address space, and the 5-level paging hierarchy doesn't change that.

Even if it were about RAM, current processors can already address far more RAM than motherboards have places to plug it in, or that anybody outside of a supercomputer center could afford.

I do not agree that the extremely specialized nature of this topic is not a problem here. I know what you are saying in that rigorously interpreted, point 3A only applies to the hook. But the whole concept will still be completely opaque to the vast majority of readers. As in "no, I didn't know that, and even after reading the article I have no idea what it meant." What good will it do the general reader to attract them to an article that they're not going to understand? Heck, most developers aren't aware of or care about this stuff any more. Jeh (talk) 22:51, 2 May 2018 (UTC)

That's not the case. I assure you that the developers are aware of this stuff. I think that given the sort of readership we have, the article will do alright at DYK. Hawkeye7 (discuss) 12:18, 3 May 2018 (UTC)

This has been stalled for two months now. A second opinion is needed on the interest of the proposed hook(s), so this can move forward. Thanks. BlueMoonset (talk) 03:05, 6 July 2018 (UTC)

@BlueMoonset: Of the three hooks, I think only ALT0 would really work and be interesting to a general audience (I don't find Pgallert's proposed ALT3 hooky), if only because readers may be curious as to what "walk further" means. It could work in the quirky slot. I'm also endorsing the previous technical review of the DYK requirements and I can see no more problems with the article. This should be good to go, but as a courtesy, I'll wait for responses from previous reviewers @Pgallert, Jeh, and Hawkeye7:. Narutolovehinata5 ^tccsdnew 09:41, 6 July 2018 (UTC)

In my view the original hook would be best. I do have an issue with the usage of terabytes / petabytes, but that's clearly a minority view. --Pgallert (talk) 09:17, 11 July 2018 (UTC)

I still think it belongs nowhere on the front page with any of the hooks. Whatever happened to "Wikipedia is written for a general audience"? That should apply to the front page more than any other place! The "walk farther" (and if you use "walk" it should be "farther", not "further") hook will lead the general audience to just a puzzled expression (huh? what? what's "walking" mean here?). And all three are misleading as phrased in that the Intel whitepaper explicitly disclaims that there are any such "plans"; the whitepaper only describes a possible future development. Show me an Intel "future processors" roadmap that mentions this and I'll shut up about that point. Jeh (talk) 18:21, 11 July 2018 (UTC)

After much discussion here, I am ready to mark this for closure. There are concerns about the accuracy of the hook, and concerns about interest to a broad audience, and given that this nomination has been ongoing since April, it appears that moving forward is proving to be difficult. I will wait for a response from the nominator here, otherwise this will be marked for closure in the next few days. Narutolovehinata5 ^tccsdnew 23:12, 11 July 2018 (UTC)
Despite many comments on this nomination page and elsewhere, there have been no major edits to the article since May. The nominator has not responded to multiple pings and talk page messages. As there are still outstanding issues that have not been addressed, unless another nominator adopts this nomination, this is now marked for closure as stale. Narutolovehinata5 ^tccsdnew 01:00, 17 July 2018 (UTC)

Yeah, I can't really address the issues, certainly not currently. ∰Bellezzasolo✡ Discuss 09:12, 18 July 2018 (UTC)