Moore's law and why persistence may not be necessary.

Ralph Johnson johnson at cs.uiuc.edu
Wed Jan 23 15:12:52 UTC 2002


	From: Scott A Crosby <crosby at qwes.math.cmu.edu>

	If you think I'm in error in this case, I'm happy to be convinced
	otherwise.

For what it is worth, I am 100% on your side.  Real memory is
good enough for 99% of applications.  I've seen lots of people
kill themselves trying to make the database fast, but the total
size of the database is only 100 meg! Sure, everything needs to
get written to disk eventually, but read it in, work on it, and
write it out at the end.  Sometimes we need a logging mechanism
for long running programs, but we usually don't need a real
database.  What is the current memory price, something like $500
per gig?  That means that $2000 fills your address space.  Why
can't a businessman put $10,000 of memory on his computer if he
wants to make his database fast? 

I'm working on a problem (in FORTRAN) that needs 50 gig of
storage.  We keep it in memory by running it on a 1000 processor
machine. That is $25,000 worth of memory for a problem that
NASA is paying several million to solve.  Memory is cheap.  We
need to be able to use it.

-Ralph Johnson






More information about the Squeak-dev mailing list