Moved blog – please update your links/subscriptions

I have moved this blog to Please update your links/subscriptions. RSS for new address is here.


Semiconductors economics

There is a really interesting article the IEEE Spectrum by Clayton M. Christensen, Steven King, Matt Verlinden, and Woodward Yang. It basically tells the story of them apply the TPS (Toyota Production System) to semiconductor manufacturing and results of doing this.  Pretty amazing results.  The most amazing change is that this methodology allows smaller runs of chips while still being profitable and producing the chips at competitive prices.  This is a game changer in silicon as it allows silicon to more closely match the trajectory of ever-shorter product life cycles in consumer electronics.

Who wants to guess the company that is the subject of the article (purposely withheld in the article)?

Content goes from king to the new loss leader

By now it’s fairly well-known that selling digital music downloads is a tough business with thin margins.  However, across both online as well as brick-and-mortar retailing, music and video is being used more universally as a loss leader and it’s not clear what this trends means for content in general.  For example:

  • for Apple, digital downloads are not something they make a lot of money on, however, it enables their hardware universe (iPhone, iPod, Apple TV).
  • for Amazon, many people have suggested that digital downloads is a traditional retailing loss leader.  It drives people to their page, where Amazon hopes to sell them something above and beyond the digital download.
  • for wal-mart, CDs and DVDs have been used for a long time as a loss leader.

The New York Times is reporting that Apple’s new deal to sell movie downloads (as opposed to their rental business) is not a great deal for Apple from a dollar perspective.  So, Apple appears to be using the same model here as they did with the iPod.  Effectively they are relegating the movie downloads to a loss leader in the hopes of selling more Apple TVs.

So is content no longer king?

Started another blog on Tumblr

Started a tumblog.  Check it out.

Live Mesh – the ultimate P2P platform

Live Mesh was introduced yesterday by Microsoft.  It is the brain child of Ray Ozzie and really seems like an out-growth of the work he did with Groove.  At its simplest, Live Mesh is basic file sync.  But another way to think about Live Mesh is as the ultimate P2P platform.  When people think of P2P they think about Kazaa, Napster etc where everyone’s files are available to everyone else.  Live Mesh is different from vanilla P2P because it adds the following:

  • control over who sees what files
  • creates a personal P2P cloud amongst your own devices (file sync)
  • the concept of your own cloud store
  • a platform, which will let ISVs build applications on top of it

So what does this mean?  Live Mesh has the following implications:

  1. Cloud storage companies can kiss their businesses good bye because cloud storage is integrated with Live Mesh.  Cloud storage is a thin margin business and only those with scale can make it work.  After the shake-out likely survivors are MSFT, GOOG and AMZN.  I should be clear that there may be still be a lot of 3rd party applications that leverage the infrastructure of the big 3 but if you are doing your own storage thing, you better switch now.
  2. Sync is a big deal because if done right it has the ability to transform our digital lives.  I know that sounds like hyperbole but if you have more than one computer you will likely have blown productivity managing your files.  It basically makes computers (and devices) more useful.  There is a lot of people playing in this space but by virtue of creating a platform and a standard set of APIs and protocols, MSFT has the chance to win big time.
  3. MSFT has an advantage over GOOG and AMZN here because they have better distribution (by virtue of their operating system).  Additionally, they have been marketing to ISVs for years.  Don’t underestimate this.  Once Live Mesh is integrated into Visual Studio, any kid with rudimentary Basic skills will be able to build applications on top of the mesh.
  4. Going back to the P2P angle, Live Mesh can be a content delivery platform.  If I were a content company I would start thinking about this right now.  Streaming media has economic issues that Live Mesh solves.  Microsoft should be out licensing this to every device maker on earth.  The consumer electronics companies will be reluctant participants but they should embrace it because it enables them to compete with Apple and frankly they need help in this area.  If you look at Ray Ozzie’s memo, he talks about the 3Cs the first being content.
  5. Social networks are about interaction whether it be IM, posting or sharing.  Guess what?  Live Mesh addresses posting and sharing and if done right will make it easier.  YouTube flourished because it let people embed videos in HTML.  Well, Live Mesh can take this to the next level, facilitating what Fred Wilson calls micro chunking.

However, the biggest key to Live Mesh’s success will be how much lock-in is baked into the platform.  If all protocols are published and any one can interact with Live Mesh completely at the protocol level and for free, then Live Mesh could be transformative.  If the above is not true, the Live Mesh will fail because lock-in does not work on the web.

Shocker – Apple buys PA Semi

I am not sure what to make of this but Apple has apparently bought PA Semi for $278M.  When Apple made the switch to Intel, PA Semi was supposedly in the mix, so I am not sure what it means that they have now acquired  PA Semi.  Apparently, it will give them differentiation in the iPhone space.  While vertical integration has arguably been successful for Apple, this seems like a bridge too far although it is hard to question Apple at this point.  They sure know how to change the game.  Perhaps with PA Semi’s processor Apple will have to horsepower to run everything in software (baseband, application processor, video decoding) on the iPhone.  This is the holy grail but people have been saying this is not realistic.

For the investors, this is likely not that great of an exit as a I believe a lot of money went into PA Semi and the price leads me to believe that this was somewhat of a bargain for Apple (considering what investors were looking for).  You have to believe Intel is the odd man out in this.  I wonder if Apple was unhappy about something Intel was doing (or not doing)?  TI is an investor in PA Semi but it could spell trouble for them as well if the iPhone market share grows and PA Semi manages to change the game in mobile phone silicon while the iPhone market share grows.  Apple always has have had a cozy relationship with their semiconductor partners including IBM with the PowerPC so from that perspective it is not unprecedented.  I am guessing that PA Semi chips will not be in the 3G iPhone but perhaps the generation afterwards.

Yahoo earnings – status quo, eh?

So Yahoo earnings are out and they are a little better than expected.  But you have to think Yahoo tried very hard to beat the expectations pulling all possible levers to make this happen.  If that assumption is true then the real numbers are probably more inline with what the street expected (or worse).  So now what?  Well, Ballmer said he doesn’t care what the earnings are.  That makes sense because if MSFT did their homework, they already figured out what Yahoo was worth to them.  I guess I don’t understand how Yahoo could argue that they are worth so much more.  The stock was sub $20 before the bid was launched so in what reality is it worth $40 (or whatever they want).  Hipmojo does a nice analysis here.