Moore’s Law is really Moore Effect – Lessons for Apple iPod ?

 

Background

Moore’s Law states that “Semiconductor Technology doubles its effectiveness every 18 months”.  Now if we look at other areas of  computer technology, we seem to be seeing a similar rate of tremendous progress though it may not be exponential as in the case of  Semiconductor Technology. This kind of  progress  is due to what I call as the Moore Effect – “a phenomenon that enables a particular technology to improve at a rate similar to that proposed by Moore’s Law”  as Moore’s Law may not be applicable to all technologies”.  Based on Systems Thinking philosophy, we can call Moore Effect as a virtuous cycle of the ultimate kind.

 
As I probed further, I realized that Moore Effect would kick in whenever a  large number of people are working on a technology. This gave raise to the question – why would a large number of people work on a particular technology ?

In the September 1998 issue of IEEE Spectrum, Robert Lucky had written an extremely thought provoking and interesting article on Moore’s Law.  In that article, he had proposed an interesting theory which he called “fashion theory”. According to that theory,  many people start  to work on a technology because it is fashionable and progress is directly related to the number of people working on a particular technology. 
 
I had a strong feeling that there must be a better explanation for this. As I probed further, I realized that the answer lies in the fundamental business principle of  competitive advantage – that something which helps a company or an individual keep ahead of or stay in-step with the competition. Therefore, companies/individuals adopt a technology when they spot the potential to leapfrog competition or to create new markets or simply to derive benefits that their competitors have derived.  When we find more people working on something, it should actually be more people creating applications or add-ons or other value adding activities which put pressure on the originators of the technology to improve further to meet the growing demands of the people working on the technology.  IMHO,  Semiconductor Technology is  advancing fast mainly because of the extraordinary demands placed by the ever expanding use of semiconductors (“applications”).

In the subsequent sections, we will see some key examples of how Moore Effect in the technology industry.

Moore Effect in Personal Computers

 
It is clear to everyone that the PC is the winner of the desktop battle and Apple is the clear loser, though Apple seems to have re-invented its magic with the iPod, iMac, iBook etc. When IBM decided to build the PC, it chose to keep the specifications open,  so that other companies could also build PCs. IBM did this because they woke up late to the PC phenomenon in 1980 and they wanted put a IBM PC on the market within 12 months. So they chose to build the PC using third party components resulting in the open architecture of the IBM PC. The open architecture helped the PC industry dominate the computer industry and also resulted in the unprecedented dominance of the industry by Microsoft and Intel – the Wintel duopoly.  Due to the immense penetration of the PC in the office as well as the home markets, the PC industry offers tremendous business potential for any company to take advantage of. This has led multitudes of companies to enter the PC industry. This has resulted in the creation of newer and newer, often times competing technologies that are improving the PC – witness the growth in the number of types of PC brands, types of peripherals, types of software, special technologies such as speech recognition etc.

On the other hand, none of this intense competition is seen in the Apple segment because Apple chose to keep the specifications closed thereby eliminating the potential for competitive advantage for third parties. This also means that Apple as a single entity has to compete against the likes of Microsoft, Intel, Dell, IBM, HP/Compaq all of whom are constantly working to improve PC technology.  Further proof of this lies in Apple embracing PC standards such as USB in its iMac, iBook and PowerMac line, which has enabled Apple to take advantage of the advances made by the PC Peripheral industry.

 

As per the technology forecasts for the PC industry, PC installed base is around 90%. In contrast, the Mac installed base has dwindled to 3%. Incidentally, it was Apple II which was introduced in 1977 that set off the PC industry revolution. IBM PC was introduced only in 1981 with the MSDOS  operating system.

Microsoft Windows has won the operating system wars much to the chagrin of IBM which tried in vain to push OS/2 as the dominant PC operating system. Microsoft won the battle because it had convinced scores of third party developers to develop for the Windows operating system resulting in a flood of software designed to run on Windows. Since there were more applications available for Windows, it became amply clear to businesses and home users that they should buy PCs with Windows operating system pre-loaded.

Thus Microsoft triggered the Moore Effect and ensured that it will continue to dominate the industry due to its sheer volume of installed base.

 

Moore Effect in Open Source

When Linus Torvalds created the Linux operating system and published the source code, little did he realize that he was going to create a gargantuan development community that has spawned so many great open source projects Apache, Struts, Tomcat, Mozilla that have literally changed the world. In fact, Open Source has taken Moore Effect to the next  level by engaging numerous independent programmers and big corporations like IBM, Novell etc.

Another interesting exmple from the history of computing is the Unix operating system. Unix was developed in 1970 by Dennis Ritchie and Kenneth Thomson. AT & T  adopted an interesting tactic that made Unix popular – it gave away the source code of Unix to many universities which ensured that scores of academic researchers worked on Unix and helped in advancing Unix. AT&T also gave the source code to third party companies which resulted in the many commercial versions of Unix  such as HP-UX, IBM AIX, Sun Solaris, SCO Unix/Xenix which further enhanced the popularity of Unix.

How do we go about triggering a Moore Effect

 

If we closely analyze the instances of Moore Effect mentioned above, it will not be difficult to infer that a platform has to exist, most likely based on open standards that adds value to end users and at the same time allows third parties to create additional value. The industry will always try to move towards non-proprietary standards but there will still be standards that are proprietary like the Microsoft Windows platform. Some innovative companies will create new markets with some proprietary standards, which might become non-proprietary over a period of time.

 
An interesting example is the cell phone browser technology called WAP developed by a company called Unwired Planet (now Openwave), which has become a de-facto standard. WAP has become somewhat of a bad egg because of the cumbersome user interface and slow speeds . But as an example of Moore Effect, it makes for a good case study. The company gave away critical pieces of its technology to customers and competitors to speed adoption. By trading away its monopoly, the company made WAP a standard feature on mobile phones and boost the market for Unwired Planet software. The company’s marketing director Benjamin Linder said “Its much better to have a 50 percent share of a market that’s 10 times as large, as any MBA will tell you”.  

Moore Effect gets built by two important drivers – End User Adoption Rate and Third Party Application Creation Rate. End users start adopting a technology when they see competitive advantage or value in using that technology. As End User Adoption Rate increases, the market starts to grow which attracts the third parties that see competitive advantage in enhancing or adding on to the platform technology. Since the third parties are actively involved in advancing and promoting the technology, End User Adoption Rate increases further bringing in more third parties which pushes the Third Party Application Creation Rate further along and so on.   

 

Conclusion


Metcalfe’s Law doesn’t cut it (Point Counterpoint).  It may explain why a fax or a telephone network becomes more useful as the number of phones and faxes increases, but the Internet is a whole other planet.  I would argue that the reason the Internet has rates of adoption that is orders of magnitude better than radio or television is due to the fact that Third Party Application Creation Rate (rate of creation of websites)  was extremely high as almost anyone could cobble together a website quickly and as the number of websites grew and made the Internet more valuable, the number of Internet users kept growing in even greater proportion.  IMHO, Blogs will further increase the Internet End User Adoption Rate because potentially every end user can become a Blogger (Third Party Application Creator).  Going by this argument it is easy to see why Radio or Television took a long time, because it takes a lot of investment to create a Radio or Television Station (Third Party Application Creators).  Even  creating a  Radio or  Television program is expensive as you need access to studio quality equipment and expesnive people who can operate the equipment.

It is amply clear from the above that companies aiming for unsurpassed dominance in the technology industry should focus on creating  a Moore Effect around their products. Most recent example is Apple’s iPod which has captured the imagination of scores of  people, third party companies, record labels, artists etc. Although several third party hardware add-ons for the iPod have been released, the real power is in the software. By publishing an API for the software platform in the iPod and making iTunes into a platform with APIs for Plug-ins, Apple could take iPod/iTunes to the next level as it would allow independent developers to create applications for the iPod/iTunes that Apple may not have thought of or does not have the resources to develop.  Will Apple do it ? It seems to be moving in this general direction with the iTunes COM component.
Others on the web talking about this:
1.  Apple Please Open up the  iPod API
2. Ipod, therefore Iam

Bibliography (not complete)

1.      IEEE Spectrum, September 1998 issue, article titled “Reflections” by Robert W. Lucky p17.

2.      IEEE Computer, October 1996 issue, various articles on “50 years of computing”.

3.      IEEE Spectrum, January 1997 issue, various articles on “Technology Forecasts”.

4.      St. Petersburg Times, July 6 , 1998 issue, article titled “C-phones roam the Web”.

5.      The Fifth Discipline by Peter Senge

6.     Story of Apple by  Jim Carlton

Comments

  1. Anonymous said August 18, 2004, 7:22 am:

    I have to disagree with a large part of this seemingly “PC is wonderful and so open” Article. Apple I do believe have pushed the standards by being the first to take on standards such as USB, Firewire, Bluetooth, Wireless.. I could go on. It is the PC market that have had to keep catching up with what Apple have been putting in their hardware & even software (MS Windows & Longhorn with their expose rip off amongst others)

    Windows, Open? I dont think so. Apple, Open? Well ok, not completley but by taking on more open technologies and standards they are definitley more open (See Darwin)

  2. Anonymous said August 19, 2004, 12:24 am:

    Dear Anonymous,

    What i mention as the PC’s open architecture is due to historical reasons. i am sure IBM did not do that with an altruistic goal., it was just the result of their need to respond quickly to the threat posed by the then Apple without thinking through the strategic implications of their decision which handed control of the industry to the Wintel duopoly.

    You are right in that Apple is now fully behind standards and did introduce firewire (tried to charge for it, but failed !), 802.11b and now 802.11g, bluetooth etc. Apple’s support for standards is truly laudable.

    However, you are wrong when you include USB in the list of Apple firsts. USB was invented by Intel and again it did try to develop it in a proprietary way and we had 2-3 different specifications at that time but better sense prevailed and it got unified. Apple was not the first to introduce USB. As i said in my article, Apple was smart enough to embrace it quickly (in the iMac) which resulted in its ability to tap into the PC peripheral industries’s prolific output. iMac was introduced in 1998 whereas PCs had USB support in 1996/1997 (Windows 95 osr2 released early 1997 had USB support builtin).

    If you want a history lesson ion USB, i strongly recommend the following links:

    http://www.mindshare.com/pdf/usb10to11.doc

    http://www.eetimes.com/article/showArticle.jhtml?articleId=18305126&sub_taxonomyID=2517

  3. Anonymous said August 20, 2004, 2:08 am:

    Geoffrey Moores “Inside the Tornado” has some good thoughts as to why Apple could not have become another Microsoft, and I would tend to agree with him. It is very interesting read.

    One of the things that MS had going for it was DOS, and IMHO, one of the very smartest things they did was to maintain backwards compatibilty between subsequent releases of their product. There was even an interview with an ex MS employee about how MS bent backwards (not fixing bugs that became a feature) to support backward compatiblity.

  4. Anonymous said August 20, 2004, 12:06 pm:

    You are right Ganesh. MS is one of the best at “parlay”-ing.