Calculating M$’s Attach Rate for That Other OS

Nowhere is this information published so M$ must be suppressing it. However, we have the tools, simple arithmetic.


  • installed base of PCs is about 1500 million
  • with death and new PC production, the number of PCs alive grows by about 100 million annually, 360 million new PCs and 260 million discarded/dead PCs
  • share of PCs running that other OS drops 2% per annum

Using NetApplications latest numbers, M$’s share a year ago was about 93.78% and now is 91.86%. So the number of PCs running that other OS was 0.9378 X 1400 million = 1312.9 million. Now it is 0.9186 X 1500 million = 1377.9. Of the 260 million deaths, about (93.78+91.86)/2 percent ran that other OS = 241 million.
1312.9 million – 241 million + x X 360 million = 1377.9 million
x = (1377.9 – 1312.9 + 241)/360 = 85%
That means of the 360 million new PCs produced/sold in the past year, only 85% run that other OS. If we use M$’s number of 50million per quarter, 200 million for the year, the share of units shipped is even smaller, (1377.9 – 1312.9 + 200)/360 = 73.6%. The difference could be the illegal copies we hear about.

We know Apple shipped 16.7 million in the last year so the GNU/Linux units are likely (360 – 360*.85 – 16.7 = 37.3 million)/360 = 10.3%. They are selling somewhere.

This entry was posted in technology. Bookmark the permalink.

32 Responses to Calculating M$’s Attach Rate for That Other OS

  1. oiaohm says:

    Clarence Moon you are a goof ball. It does not make much research to find out you have been lied to.

    When it comes to databases there is not black and white answer if your engineers are saying this “Our engineers say that it is superior product ” They should be fired on the spot.

    Apache Derby, Ingres, Mysql and Postgresql all support unlimited size database. MS SQL don’t support this.

    Each has there own unique advantages over MS SQL and unique disadvantages to MS SQL.

    So database selection should be a task at hand selection based on workload requirements. Yes most cases the workload is better served by 1 of the 4 than MS SQL servers.

    SqlServer Express Edition. is crap on that table of compare feature it has is the same as Microsoft SQL Server Compact (Embedded Database).

  2. Dr Loser says:

    “Clarence Moon?”

    Good Lord.

  3. Conzo says:

    Nice stuff! I know accelerators only from the end-user side (x-rays at Hasylab and Argonne, neutrons at NIST and PSI Switzerland), and developed a rapid dislike for doing measurements (way too much ‘measure the blanks in the book with tables of whatever you’re studying), but I’ve always found the actual technology very interesting – both with regard to the design and engineering of the facility and of the experimental equipment. In any case, I think it’s a lot better job than the usual measurement slavework 🙂

  4. pogson says:

    I had diverse roles but my studies were in nuclear instrumentation and my work at the Cyclotron Laboratory were related the the particle accelerator. I helped design the nuclear spin filter ion source by simulating the spin filter and beam transport. I designed an improved beam buncher and the control system for the magnetic field mapping system. I analyzed the magnetic field and did CAD for the central region of the existing cyclotron and the proposed new cyclotron.

  5. Conzo says:

    @Pogs: Thanks! It’d be interesting to leaf through it, but yeah 80s and before didn’t really feature PDF and online archives. Don’t break a sweat over it 🙂

    I’m following the OS wars scene a bit from a distance (with no mission except to rehabilitate the penguin from its association with Linux), and it interests me whenever I get the chance to get some insight into who the people in both camps are.

    Did you do measure at the synchrotron, or was it just that your thesis was mentioned in the report?

  6. Clarence Moon says:

    We do not use MySQL in our products. We do use the SqlServer Express Edition. Our engineers say that it is superior product and they did evaluate MySQL. On what basis do you claim that MySQL is better?

  7. Ray says:

    assumptions are only theoritical.

  8. Clarence Moon says:

    It is some sort of South American dance step, IIRC.

  9. Kozmcrae says:

    I guess Clarence Moon has never heard of Samba. Either that or his memory challenged synapses don’t reach that far.

    You don’t get around much, do you Clarence.

  10. pogson says:

    Yes, to show that MySQL has been superior in some ways for nearly a decade. The others are from 2009.

  11. NT JERKFACE says:

    Pogson did you really just post a db benchmark from 2002?

  12. Clarence Moon says:

    Microsoft has been fined in Europe for violations mainly involving packaging of their MP3 player and browser and were forced to sell Windows without a media player and to let new buyers select which browser they wanted. As far as I can tell, they were not fined in the US, but they were sued for being a monopoly and blocking use of the Netscape browser.

    Honestly that doesn’t seem so terrible to me and I think the Europe courts were just after the money, which was billions of euros, I believe.

    Even so, they were found guilty, but that didn’t have anything to do with their Windows software outside of causing the monopoly in the first place. If they didn’t have the monopoly, they would not have been fined, but they were not fined for having the monopoly, but for using it for MP3 and browsers.

    I don’t think that Windows is terrible either, but it is fair to say that they have their problems and that there may be a better way. But you have to tell the users and convince them that things are so bad and that there is something to use instead. I don’t see that happening anywhere. Grass roots is wonderful sometimes, but you are talking about more than a billion computers, using your figures above.

  13. pogson says:

    Fair enough but M$ has not played fair. They have lied and had other lie for them. M$ has done illegal acts to stifle competition. They should be stopped. That will not happen if people don’t talk about what M$ does. I will not desist simply because they stayed out of jail/were not stopped.

    I have many times showed side-by-side on the same hardware GNU/Linux and that other OS. People do care that they have been paying top dollar for less performance and seen their investment in IT deteriorate over time with malware, slowing down and re-re-reboots.

  14. pogson says:

    Running “slow in Windows” is something M$ encourages. We have seen them repeatedly interfere with competition.

    Here’s a benchmark:

    “Overall, Oracle9i and MySQL had the best performance and scalability (see charts, images 1 and 2 in slideshow), with Oracle9i just very slightly ahead of MySQL for most of the run. ASE, DB2, Oracle9i and MySQL finished in a dead heat up to about 550 Web users. At this point, ASEs performance leveled off at 500 pages per second, about 100 pages per second less than Oracle9is and MySQLs leveling-off point of about 600 pages per second. DB2s performance dropped substantially, leveling off at 200 pages per second under high loads.

    Due to its significant JDBC (Java Database Connectivity) driver problems, SQL Server was limited to about 200 pages per second for the entire test.”

    see eWeek

    Of course, MySQL also holds some world-records

  15. Clarence Moon says:

    I think that it is useful for various people to promote their favorite Linux distribution, FOSS program, proprietary program, or even Windows OS. That is the way it is in the real world and vendors or advocates are free to say what they think.

    I think that it is not very useful to gratuitously denigrate who someone thinks is an opponent or what is just an opposite point of view, though. In my business, nothing happens until a prospective customer first of all hears about us and is willing to sit down and listen to our proposition. I know from experience that knocking the opposition does not work. The client probably never heard of our competitors either, so we would just be doing his advertising for him.

    I see a lot of that here. Instead it would be better to just extol the merits of whatever OS or application one is touting and be done with it. People who might read the proposals can make their own choices.

  16. pogson says:

    I am so old the thesis may not be on the web. Let’s see what Google finds:
    Here’s the annual report of the University of Manitoba Cyclotron for 1982-1983 where I was mentioned several times. That was a few years after the thesis.

    The University of Manitoba Library has a copy.

    I have a copy. I suppose I could scan it and put it up but it’s hardly Earth-shaking after all these years.

    Google knows about it but I haven’t been able to find the contents:

    The Library of Canada has a copy but apparently it’s not indexed…

  17. Conzo says:

    “I have a M.Sc. in Nuclear Physics. My Maths is OK.”

    Dissertation, or it didn’t happen!

    (no, honestly, I’m curious)

    Every time someone installs a Linux, a penguin facepalms …

  18. NT JERKFACE says:

    Pogson sez:
    NT JERFACE wrote a bunch of silly things. I have a M.Sc. in Nuclear Physics. My Maths is OK.

    Your matheses is ok? Maybe you can multiply but your data sucks. Net Applications doesn’t report the installed base, just what is being used to surf the internet. When we say Linux has 1% we are actually being generous. Net Applications can’t measure how many office or grammy machines exist that rarely connect to the internet.

    Ohio sez:
    SQL Server sales are mostly vendor lockin. Ie they made databases in access now need it larger.

    OHIO please stick to what you know. Extracting data from access is easy:

    This is the type of denial I am talking about. You can’t accept the possibility that a company might buy from M$ for reasons other than lock-in/inertia/conspiracy/extortion/satanism.

    MySQL runs SLLLOOOOWWWW in Windows and even the devs admitted to it at one point. It does not have as many features as SQL Server because *GASP* M$ used their $$$$$ to add features that Windows devs were requesting. $$$$$$ can buy developers to improve ANY software, including ebil M$ware. You guys are in serious denial of this basic fact.

    If I pointed out that the open source world has nothing on Skyrim you would agree. Bethesda clearly has the cash to hire game developers that are not willing to work 40 hours a week for free. But when it comes to something like database software you have this preconceived notion of anything M$ makes being inferior since it is an area where open source has been competitive.

  19. pogson says:

    It was not a free market for at least a decade. It is coming unstuck now. M$ still does not deserve the share it has. I expect ultimately M$ will have less than 50% share in a few years. The world wants to progress faster than M$ can follow. Ideally they would be banned from doing business but that does not seem to be on governments’ agenda.

  20. Clarence Moon says:

    I don’t see where it is so bad that Microsoft does not have all of the world’s business. If they had 100% of that business, they would not have anywhere to grow. If they used to have more, then they have room to improve their game and the incentive to do just that, I think.

    The old saying is “You can’t win them all!” and I think that it applies to Microsoft as well as the Yankees. Apple grows by advertising their wares just as Google has advertised Android along with Samsung, HTC, and others and grows the Android business.

    Microsoft advertises, too, and they will just have to be satisfied with their 20 billion bucks worth of the action. The fact that they are not getting all the business just shows that a free market is at work.

  21. Kozmcrae says:

    Phenom is practicing to be a literary contortionist. He got 30 seconds out of me. Not a bad start. What is it about you and communists anyway? Did one beat you up at recess?

  22. oiaohm says:

    NT JERKFACE maybe Mysql is the walking dead we are not 100 percent sure yet. This might explain it lack of recent feature growth.

    mariadb is mysql old lead developer.

    SQL Server sales are mostly vendor lockin. Ie they made databases in access now need it larger.

    Taking sweet time with samba 4 is partly because samba 3 does offer something Windows 2000 and before does not. Even something modern 2008 does not.

    Really its a trick of ldap. Means to have servers in different locations with the same login information and appear to machine to be the exact same computer due to samba not respecting system id. Basically network isolation is also possible.

    Now where is also live cluster install option in Samba 4 as well. Yes where many machines appear to be one servicing the ads need in a local network.

    Samba people have not made there job simple. They are also wanting to maintain the PDC and BDC of pre windows 2000 as well. Does have its advantages for controlling fall over.

    That is the problem samba 4 is aiming to do more than Windows 2000 or 2008 and some of the old things samba 3 does that MS removed in 2000 are still useful. Samba developers really don’t seam to like the easy road.

  23. pogson says:

    NT JERFACE wrote a bunch of silly things. I have a M.Sc. in Nuclear Physics. My Maths is OK.

    The 1% number is the myth, not that GNU/Linux is doing wonderfully well without much promotion. Apple sells into a niche. Schools used to buy that PCs were complicated and that Apple knew best but then Apple started using x86, too. GNU/Linux share is greater than Apple’s by several measures:, 5%, and Wikipedia, 3.48%. The question you should be asking is “What’s wrong with NetApplications?” and “How can Apple be that high when it publishes much lower unit sales and has only a few stores in China?”. Apple doesn’t even try to sell globally, especially in emerging markets where there are first-time buyers with not a lot of cash.

    M$ produces such a whimpy OS it cannot live in any number without M$’s servers. Same with Independent Software Vendors who are not independent but encouraged and rewarded for using M$’s services. “ISV”s are actually paid to include M$’s products in their software packages. It’s not the end-user who is choosing all those database management systems from M$, but the ISV and the ISV is not unbiased.

  24. NT JERKFACE says:

    I’m not going to comment on your math because I don’t care enough.

    I think the real question is why Linux has 1% while Apple’s share keeps increasing.

    M$ is the big bugaboo but Apple can somehow increase the sales of their non-Windows computers that have a high markup?

    Forget the desktop, how does ebil M$ keep making billions from Windows Server? How are they able to sell SQL Server when MySQL is free? MySQL installs easily on Windows Server and plugs into .NET. You think most Windows devs are just stoooped and like to throw money away? Or perhaps M$ continues to add features that Windows devs find useful while MySQL for Windows stagnates and is ignored by the people’s programmer army. That same people’s army is taking their sweet time with Samba 4 while Windows Server continues to erase reasons to switch to Linux.

    The entire open source world is in deep denial of its limitations. You guys think Linux won the server market and yet Linux still can’t offer all the features of Windows 2000.

  25. pogson says:

    M$ used to get it’s OS on 95%+ of PCs shipped by OEMs. They are way down from that number today because OEMs ship FreeDOS, GNU/Linux, and no OS. While the number of PCs shipped has relentlessly increased, M$’s share of them has dramatically decreased.

  26. Clarence Moon says:

    I do not understand just what the controversy is here. The references sort of show that Windows revenues increase from 15 billion dollars in 2009 to 20 billion dollars in 2011. That may not be as startling of a change as it used to be, but there has been a worldwide recession in effect and some increase is a lot better than the decreases that many industries, including my own, have experienced.

    There are a lot of numbers in those references, so I may have missed the gist of it all.

  27. pogson says:

    M$ cares. They talk about “attach rate” but never give a number…

    “Windows attach rate. This factor contributed to 1% of the overall 14% OEM revenue drop in FY’09 and refers to the number of PCs with a copy of Windows that OEMs purchased during the quarter. Two factors influence attach rate: piracy and OEM inventory fluctuations. (OEMs sometimes stock up on Windows one quarter ahead of anticipated price changes, but don’t necessarily ship all the copies they purchased until future quarters.) Ballmer said that the 1% drop was caused mostly by inventory fluctuations, over which Microsoft has limited control.”

    Coy, eh?

  28. pogson says:

    Clarence Moon wrote, “Microsoft and Apple’s businesses are still growing in terms of both volume and revenues.”

    For the client OS, my main concern, M$ is decreasing in share, units shipped, $ income (revenue – costs of revenue), all measures. That has been going on for several quarters now. e.g. while PCs shipped globally increased 3.6%, M$’s revenue increased by less and income was flat.

    3 months ending Sept. 30
    2011 revenue $4832 million 2010 revenue $4705 up 2.7%

    2011 income $3219 million 2010 income $3210 up 0.28%

    PC shipments were up 3.6%. So, little by little, M$ is losing its grip. Revenue/PC = $4832 million/91.879 million = $52.59. Income per PC = $35. can get $27 per PC shipping Debian GNU/Linux. In the good old days, M$ boasted of getting revenue of $2485/42 per PC = $59/PC and income per PC = $47.

    see 10Q Q1 2003

  29. Clarence Moon says:

    When it comes to business, the first concern is how well the business is doing. If the business is not doing as well as expected, then it is necessary to look for reasons why that may be.

    Microsoft and Apple’s businesses are still growing in terms of both volume and revenues. Apple is growing a lot faster lately and the new success of the Mac is pretty strong evidence that they are successfully carving out a nice business in the upper end of the laptop range.

    I don’s see where it is possible for Microsoft to publish any information here outher than numbers pertinent to Windows sales. If such data was necessary, then statistics companies like Gartner would come up with such data to meet the needs of those who might need to use it. When Gartner and other publish such data, it seems to always show Windows and Apple as taking close to 100% of the business.

    Statistics from internet usage only really describe internet usage and trying to make them tell a story about unit sales is perilous and likely to come up with a wrong answer since the sales and the usage in one specific mode are not that firmly connected.

  30. Phenom says:

    Quote: Nowhere is this information published so M$ must be suppressing it.

    Pogson, that reminds me of an old joke from an ex-communist country (Eastern Europe). The joke was about the most prominent country and party leader during the years of communism. A bit of additional info: that leader was claiming to had had a profound past as an underground agent before communists took the power, but rumors had it he was merely a simple nobody of no importance whatsoever. So, the joke went like this:

    “He was semi-underground agent: he was hiding from the police, but the police was not looking for him”.

    Your paranoia is exactly the same. No one is publishing this info, because no one cares.

  31. Kozmcrae says:

    You are making too much sense Mr. Pogson. Expect to be severely punished for that by the Microsoft gutter rats.

Leave a Reply