[Beowulf] First 96-Node Transmeta Desktop Cluster Ships

Andrew Piskorski atp at piskorski.com
Thu May 5 04:18:29 PDT 2005


On Thu, May 05, 2005 at 12:35:14AM +0200, Vincent Diepeveen wrote:

> Please note that nowadays your salaries for programmers, world wide seen, 
> are not true at all.

If I recall correctly, the example given was a "structural analyst"
costing an employer 250k USD/year; he said nothing specific about
programmers, or software.  This is a reasonable example.  There are
many, many highly skilled workers in different niches (including
programmers) around the world who cost their employers easily that,
sometimes much, much more (think Wall Street traders.)

I'm somewhat suspicious of those 96 Transmeta processors myself.  As I
think has been discussed before on this list, I suspect that a box of
Opterons might make more sense in many cases.  Which means that
Orion's obvious competitor in this new niche market might be
Rocketcalc:

  http://www.rocketcalc.com/
  http://www.orionmulti.com/

But claiming that there is no plausible market for a $100k "cluster in
a box under my desk" is just plain silly.  There most certainly is a
plausible market, the open question is how much it's worth and whether
companies like Orion can capture it.

> Secondly in this country, i would really be HAPPY with a contract job to
> parallellize software (or whatever algorithmic software) around $30k.

Vincent, in that case, I feel fairly safe in assuming that at least
one of the following must be true:

1. The economy in the Netherlands is much, much worse than I imagined.
2. You are not nearly as good a programmer as you seem to think.
3. You are very confused about the economic opportunities that are or
   are not available to you.
4. You are extremely picky in what jobs you will take, or otherwise
   have an unusually low interest in maximising your personal income.

The common theme here is that YOU personally are in no way a
representative sample of Orion's potential customers.

Although I have no idea what percentage of the US programming
workforce they might represent, I have every confidence that there are
large numbers of downright INCOMPETENT programmers in the USA making
salaries of $70k+/year (with an actual cost to their employer of
perhaps twice that), which is somewhere in the same ballpark as the
average salary for US programmers in general, depending on what survey
you believe.  And I would be astonished if European companies are much
more efficient in wringing value out of their employees than American
ones.

So although there is very wide variation, an above average programmer
should, over time, have little difficulty in grossing $70k USD or more
in salary per year, if he is aware of and pursues that goal.  (Other
countries would be somewhat different, of course.)

As for that wide variation, well, the market for programming talent
appears to be pretty innefficient.  Paul Graham has good ideas about
why this is probably so:

  http://www.paulgraham.com/gh.html

-- 
Andrew Piskorski <atp at piskorski.com>
http://www.piskorski.com/



More information about the Beowulf mailing list