Not small as in Lilliputian or pygmy, or small as in weak
or puny, but, well, tiny, and shrinking all the time. E. F. Schumacher
popularized the notion, and the term, in his landmark 1971 essays http://en.wikipedia.org/wiki/Small_Is_Beautiful),
and although his original thesis is long forgotten, the term isn’t; and so
today I am applying it to the handheld and PC markets.
These markets are shrinking in the feature size, the process
used, to make the products they sell. And yet, the parts don’t get smaller—to
the contrary, they get larger and draw more power. It’s the mystery of the
industry, some strange perversion of the law of entropy. Sometimes they shrink
and stay the same size, but they almost always end up using more power in doing
that. And since using more power results in more heat dissipation, and there’s
only one way to dissipate heat, which becomes a function of square meters of
surface area, the problem is compounded. In this case smaller is not more
beautifuler, it’s more hotter. It would take Superman to blow on it to cool it
off.
And although this tiny industry is as obsessed with size as
a New York model or a California divorcée, in one param-eter, money, it is
really huge and getting huge’er. Again, the California divorcée comes to mind:
tiny and rich.
These tiny things are in the computer market, which is
estimated to be worth $220 billion in 2005, and in the mobile phone market,
which is estimated to be worth $68 billion in 2005. It’s interesting how
excited everyone gets about the mobile phone market, and it’s not even
one-third the value of the PC market, and even less than that for the
multimedia processor suppliers. But the unit number is what gets everyone excited,
billions of units, billions! (As an
exercise for the reader, do a web search and see if you can find the market
value for the mobile phone market, worldwide in dollars. Send your results to [email protected].)
And the mobile phone market is also getting smaller, not
only in terms of the parts used, but in the number of parts suppliers. In this
issue (see p. 1) we tell the story about how ARM of Cambridge, England, has
acquired Falanx of Norway, which of course follows the spate of recent
acquisitions of smaller IP companies by larger ones, and a consolidation we
predicted in our Multimedia Devices for Handhelds market study (and yes, we are working on an update—these things take
time y’know).
And although it’s not really a consolidation, more of a
lateral shift, we also tell the story of Santa Clara–based Marvell’s
acquisition of the Intel XScale and wireless product line (begins p. 1). Will
there be more? Of course. Will all the small suppliers disappear? Yes. Will the
market be dominated by two, maybe three, big suppliers, as is the PC market?
What do you think? Of course. Soon? Before the end of this decade. (That ought
to set the lemmings off—move away from the edge or you’ll get trampled.)
The consolidation of the handheld market suppliers is
especially true when you take into consideration that the semiconductor market
for application processors in 2005 was just $21 billion, and even with all its
SoC’ness will only reach $40 billion by 2010. Now don’t get me wrong, a $40
billion market is nothing to sneeze at, but it’s not big enough to support 20
or so vendors. So the number of suppliers will get smaller. And, as we all
know, small is beautiful, right?