help with pc gene analysis

James P. H. Fuller jim at crom2.rn.com
Sat Aug 3 11:57:28 EST 1991


Peter Hakenberg <NETADM at DBNMEB1.BITNET> writes:

> You must take a old SUN and a brandnew 486 if you want the same power.
> For example a Sparcstation2 is about 4 to 5 times faster than a 386 with 33
> MHZ.

     You've got to specify the task you want to perform.  For instance, most
of the operations performed on the large biological databases like EMBL and
GenBank are disk-bound, not cpu-bound, so that the cost-effective way to do
tasks like these is to use a cheap generic 386 or 486 AT-clone and put your
money into a big, fast disk.


> Also you have to add the software-costs to the PC.

     There is an astonishing variety of high-quality biological software
that is free.  Fasta, which is used by both GenBank and FLAT DB for their
network servers, is free (large public thank-you to Messieurs Pearson and
Lipman.)  The massive and excellent GNU Unix software collection is free.
X Windows for the 386/486 is free (large public thank you to Thomas Roell.)
The software I use for reading Usenet and Bionet was free (thank you Henry,
Geoff, Kim.)
     The purpose of this system is to explore and document, for the use of
underfinanced institutions like small colleges and secondary schools,
exactly how much computational biology can be done with absolutely rock-
bottom capital investment.  The only software I purchased for crom was Unix
itself;  fullscale Unix for 386/486 machines goes for under $1000 from 
several vendors now, and with 486-clones selling in the U.S. for under $3000
and 386-clones for $1500, that gives you room to add an absolute MONSTER
SCREAMER hard disk and still bring in a price that Sun can't (I mean won't)
touch. 


> One could buy a SUN Clone or a NeXT.

     Agreed, and more than agreed.  How I *do* hope the SPARCstation clones
are a success so that *they* come down to two or three thousand dollars.
When that happens I promise you I'll run madly out waving my checkbook.
Until that happy day, though, I've put a good deal of work into determining
that what I am doing is the most cost-effective thing I can be doing.  Re-
member, generic benchmark suites are meaningless.  One MUST closely define
one's own particular needs.
                                                 Best regards,
                                                         Jim       

 
crom2 Athens GA Public Access Unix   |  i486 AT, 16mb RAM, 600mb online
   Molecular Biology                 |  AT&T Unix System V release 3.2
   Population Biology                |  Tbit PEP 19200bps  V.32  V.42/V.42bis
   Ecological Modeling               |    admin: James P. H. Fuller
   Bionet/Usenet/cnews/nn            |    {jim,root}@crom2.rn.com




More information about the Bio-soft mailing list