Hello, my name is Christopher Chan. I'm working for a small startup co.
that will make many of the vast numbers of idle computers out there on the
internet and local area networks (LANs) available for processing large
tasks. We are currently investigating computational research methods in
biotechnology and the companies that are currently using them. I would like
to engage in a brief dialogue with anyone involved in computer-aided
genetics research. Would you be willing to discuss your computational
requirements in general terms for any of the algorithms you currently use ?
Please note: this is a serious inquiry that could greatly benefit the
genetics research industry. This is not a sales effort, nor will any
individual be added to a mailing list of any kind. Any thoughts or
contributions you may have on the subject would be greatly appreciated.
To briefly elaborate, our project will allow any computer that is
connected to the Internet to participate in solving computationally
intensive problems. There is an unused resource of many thousands of
computers out there on individual desktops, and we have designed and are
building a system to allow efficient use of this untapped resource. We do
recognize that intellectual property and security issues are of paramount
importance, and our system is designed from the outset to address security
as the most critical issue.
The following questions are a guideline for replying, but please feel
free to add (or omit) anything you desire:
1) What is the name of your company ?
2) Are there any well established algorithms you currently use (or would
consider using) that could be accelerated on a network of 100's or 1000's of
computers ? (As an example, one such technique is the protein folding
simulation that IBM's new Blue Gene computer will perform)
3) Do you currently make use of any commercial software ? Which ones ?
4) Does your company employ any proprietary algorithms that would be well
suited to these enhancements ? (no need to elaborate on details of
algorithm)
5) We are currently evaluating pilot projects for a field trial of our
system. Do you have any suggestions for practical algorithms that can make
use of many machines operating in parallel? If yes, does public domain
source code for such an algorithm exist ?
6) Are there any web sites or other Internet resources you use in your daily
efforts to help guide your research or keep track of the latest trends in
your industry ?
7) Do you have any other general comments ?
8) Would you be willing to correspond via email with us, to whatever extent
your time allows ?
