Talk:Density (computer storage)

Name change
I moved this article from "computer storage density" to "memory storage density" because I think it is a more accurate term for the subject. "Data storage density", which is currently a redirect to this article, is also better, in my opinion, so I would have no objection to moving it there. If you feel that the article belongs at "computer storage density", feel free to move it back there. We can have a formal discussion about the name using the Requested moves process, if necessary. -- Kjkolb (talk) 20:52, 19 May 2009 (UTC)

False statement
The article states: "This sets a fixed lower limit, which is why most modern hard drives bottom out around $100 US retail, and have for many years now". This is no longer true now, with HDD going for <50$ and even less. 94.230.84.33 (talk) 07:06, 8 May 2010 (UTC)

The average selling price for all drives by some OEM manufacturers recently according to public 10K's has been about $70. Furthermore shouldn't we expect further base cost and therefore price reductions, particularly if the areal density growth rate continues to be low. Given the OEM price of a DVD is less than $10 it is conceivable to expect HDDs to ultimately approach such a price; plus perhaps $5 more for a disk.

The source of the HDD historical price data is not clear but to the extent it is based upon Cost of Hard Drive Storage Space such data are likely not reliable. The reference is a tabulation mainly of advertised retail prices for computer stores in eastern Canada including sales tax and adjusted to US prices. The data are also very lumpy, none in many months and many data points in some months. It has a number of other problems such as mixing bare HDDs and external HDDs. Most disk drives do not reach the end user in this manner so it is not clear that this data series has any relevance to the overall price of HDDs or price/gigabyte trends. HDDs reach the end user thru a variety of channels, but most HDDs are sold initially to OEMs who then resell them, many but not all to an end user. Probably the most relevant prices to examine would be the average OEM price per drive and average OEM price unit capacity; the former has recently been published by OEMs but the latter is generally not available. There is pretty good data in Disk/Trend from about 1978 to 1998 but prior to that most drives were sold by the mainframe manufactures for which list price is available but there is little on the discount (IBM did not ever discount, DEC usually did). An example of a surrogate for OEM data is Farming hard drives: 2 years and $1M later wherein Backblaze a relatively large buyer of HDDs is a reasonable surrogate for the industry OEM price. It is impermissible original research to come up with a meaningful data set on HDD prices. We could do several set of discrete points from reliable sources to get reliable trends; eg IBM list prices for mainframe HDDs from 1957 thru circa 1990; OEM prices from Disk/Trend from 1978 to 1988 and something like Backblase. A lot of work and I don't have time for it now so right now I will leave the statements as dubious. If any editor has other ideas please comment. Tom94022 (talk) 19:36, 7 August 2014 (UTC)
 * Any forecast, particularly this one of $15 which is four times lower than prices during the entire period of 2007-2014 and fabricated entirely from numbers picked out of the air, is mere speculation and unfit for wikipedia. Wikipedia should not forecast price, but it should report history.


 * There is no mystery to historical aggregate HDD prices, because both of the major manufacturers report average HDD selling price. They are the best source of HDD historical price data. These history data are rock solid. Average selling price has been $45–75 since 2007, and is now $60 near the middle of that range.


 * By the way, is this article correct in stating that HDD prices were $0.035 per gigabyte in 2011 before the Thailand floods? Today's mid-2014 lowest retail consumer-grade HDD prices are no better than 2011! The elephant in this room is three years of price stagnation compared to the typical price improvement of 29% per year for five decades. The elephant goes unremarked and unnoticed by the industry's most professional observers. This article dis-serves its readers with a smokescreen of outdated misinformation, overoptimism and industry booster-ism.71.128.35.13 (talk) 20:17, 11 August 2014 (UTC)
 * Actually expert speculation is just what is appropriate in a talk page; FWIW my numbers were not picked out of the air but instead are based upon an understanding of what happened to cost and price in very similar products such as FDDs, CDs and DVDs. They all started in the $100s and went to below $10 in a relatively few years.  There really isn't a lot of difference between an HDD and a DVD (other than the higher cost of the disk) so continued cost and price reduction is likely, particularly if the AD growth is low (what else to do with all the engineers).  While the price per box has been relatively stable the GMs have been improving and of course there was Thailand so the "stability" maybe transitory.
 * The two remaining manufacturers have only been reporting aggregate HDD price per drive for about ten years, prior to that it was mystery. And even aggregate is not too useful given the different unit pricing in the various market segments.  Tom94022 (talk) 07:59, 12 August 2014 (UTC)


 * FWIW, WD cost per box only goes back to 2001, history prior to that is complex and/or mysterious. Over the period 2001 to 2014 WD price has decreased at a CAGR of -30% while cost has decreased at a CAGR of -35%.  WD is a reasonable surrogate for the industry wide low price per box since they are generally more into the desktop market than Seagate.  Using linear regression would give WD a cost base of $15 in 2021 at which point the GM would be greater than 50%, which suggests what might go wrong when u use linear regression.  The low cost and price per box was in 2011 at $37 and $45 respectively after which cost increased but GM increased much more.   Whether the industry returns to its traditional price CAGR is an open question, part of which relates to the recent increase in GM (10% more or less)which in turn may be related to the continuing reduction in number of suppliers.  FWIW WD box ASP has been declining quarterly since it peaked in their Q2 FY12.  Time will tell whether it is an elephant or not.
 * FWIW WD is now publishing "Average GB Shipped" which allows calculation of WD's $/GB which in 4Q FY14 is a new low of $0.06/GB, higher than Blackblaze paid, which among other things demonstrates the problems in talking about price of HDDs but does suggest there might not be an elephant at all. Tom94022 (talk) 21:11, 12 August 2014 (UTC)

out of date
This article seems to be out of date, in most parts the most recent data is from 2006 or 2007. --MrBurns (talk) 21:08, 4 February 2011 (UTC)

Another Name Change
This article is all about Areal Density of computer storage media. Unfortunately Areal Density already is an article. I am going to change this article's name to Areal Density (Computer Storage) unless anyone objects. Tom94022 (talk) 17:25, 3 September 2013 (UTC)

Delete?
While the article is somewhat out of date, it is fixable, so why not do so? When I get around to it, I will probably fix the HDD price section. In the meantime there is now a lot of useful price information in this talk page above Tom94022 (talk) 01:07, 13 August 2014 (UTC)


 * This article should not be fixed, because fixing a technology article is not a one-time task. Technology does not stand still like a cathedral. Labor motivates tech articles, like the researchers and faculty of a university, and editing keeps technical articles up to date. To stay alive, like a shark, the article must keep moving. This article has outlived it usefulness, and it should be pruned. This article has shown over the past three years that it is unlikely to receive or even to deserve maintenance.


 * Other articles, well-maintained ones, overlap with this subject. WP should not let barnacles accumulate on this site, just hoping that one editor delivers on his or her promise to fix up the article when they get around to it. Outdated articles inevitably lower the quality of wikipedia and dis-serve its readers. Maintenance is not a one-editor task; it takes a village or even better a university. The best that might be expected from this article is wikipedia salvaging part of it and its references. Merging with the HDD article is a possibility, if you get around to it. 71.128.35.13 (talk) 19:09, 13 August 2014 (UTC)


 * Sorry 13, but your deletion reasoning has no basis in policy as far as I can see.


 * It is not as though the old information ceases to be correct, there is just a lack of modern information. We delete things because they don't meet our criteria for inclusion, being up to date is not one of them. In time I suspect this article will get attention again and if it does not then it still has value. Chillum 19:24, 13 August 2014 (UTC)

Explanations of deletion policy aside, I'm actually surprised this article hasn't been updated. The wiki process normally ensures that pages like this eventually end up as a laundry list of WP:ASOFs that some poor sap has to come trim down every 18 months. --erachima talk 19:48, 13 August 2014 (UTC)


 * OK, I see the policy permits outdated articles. I suspect the extensive overlap with other articles diverts (rational) editors away from this article. At least the readers are warned, because the article is tagged as outdated right at the top.


 * It is naive to claim that old information does not cease to be correct. Yes this holds for mathematics, the laws of Thermodynamics and many laws of science. However, old information can cease to be correct for technology applications as the following example from this article shows: “One defining electrical property is the resistance of the wires inside the chips. As the cell size decreases, through the improvements in semiconductor fabrication that lead to Moore's Law, the resistance is reduced and less power is needed to operate the cells. This, in turn, means that less electrical current is needed for operation, and thus less time is needed to send the required amount of electrical charge into the system.” This is known as Dennard scaling of transistor size, and it's no longer really correct as of around 2007. Dennard scaling has slowed, unbeknownst to readers of this outdated article. 71.128.35.13 (talk) 20:47, 13 August 2014 (UTC)

As for why it's out of date, my own guess would be that's it's a combination of the formatting and why people show up to the page. It's pretty wall-of-texty, which discourages driveby additions, and it's an article which defines a basic term in a field, which means most of the clicks it gets are from people who read the intro to make sure they understood what a sentence meant and then go back to the page they were on. One of the highest traffic pages I've made major edits to lies in a similar vein; I last updated it 6 years ago, it gets about 6k visits a month, and there have been two sentences updated since my revision. That one, fortunately, contains little time-sensitive info. --erachima talk 21:00, 13 August 2014 (UTC)
 * So fix it?


 * Three years of "outdated" tagging show that this sofixit strategy is not a long-term viable ongoing maintenance strategy. 71.128.35.13 (talk) 21:20, 13 August 2014 (UTC)
 * Well not with that attitude. Also, that "strategy" is literally how everything on the entire encyclopedia works. --erachima talk 21:29, 13 August 2014 (UTC)

New SD card close to 1 TB/sq in "limit"
"Ultra microSDXC UHS-I card Premium Edition" A 200 GB micro SD card, which is about 1/4 of a square inch, subtracting for the contacts and assuming the electronic components don't go infentisemally close to any edge, that is basically 1 TB per one sqaure inch, though this doesn't consider depth and layering of course. For all intents and purposes, a micro SD card is negligibly thick. B137 (talk) 04:02, 2 March 2015 (UTC)
 * http://www.engadget.com/2015/03/01/sandisk-200gb-microsd-card/

Move to "Areal Density"
As well as updating this page to reflect new information (can't believe that out of date tag has been there so long), I'd like to move this page to "Areal density", as suggested by Tom94022 back in 2013. At that time, it seems that there was another article there, but it seems like it has since been deleted for copyright violation, at the moment there is nothing but a redirect to the unrelated article "Area density". "Areal density" seems a much better description of this page then "Memory storage density", especially as this article has nothing to do with RAM, which is usually what is referred to by the word "memory" in computer literature (See Computer memory). Let me know if anyone has any thoughts about this. — Preceding unsigned comment added by Nren4237 (talk • contribs) 07:25, 1 August 2015 (UTC)


 * Move carried out, if anyone disagrees with this feel free to move it back. I'd also now like to look at having "Areal density" redirect to here rather than "Area density", as Google web and book searches show this to be the dominant usage. Nren4237 (talk) 02:34, 5 August 2015 (UTC)


 * Hello! I'm fine with "Areal density (computer storage)" as the new article title, but Areal density should remain a redirect to the Area density article because those two are closely connected terms (with, of course), and that's exactly why we have "(computer storage)" as a disambigation in the new article title.  Oh, by the way article titles follow the sentence capitalization style, so  the article from "Areal Density (Computer Storage)" to "Areal density (computer storage)". &mdash; Dsimic (talk &#124; contribs) 10:20, 5 August 2015 (UTC)


 * Alright, sounds good! Thanks for fixing that for me. Nren4237 (talk) 12:10, 5 August 2015 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 1 one external link on Areal density (computer storage). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20151222082741/http://www8.hp.com/h20195/V2/getpdf.aspx/4AA4-4613ENW.pdf?ver=1.0 to http://www8.hp.com/h20195/V2/getpdf.aspx/4AA4-4613ENW.pdf?ver=1.0

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at ).

Cheers.— InternetArchiveBot  (Report bug) 13:04, 17 October 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Areal density (computer storage). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20140728221058/http://ns1758.ca/winch/winchest.html to http://ns1758.ca/winch/winchest.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 19:51, 8 July 2017 (UTC)