Date: Fri, 14 Dec 2001 15:03:12 -0500 (EST) From: "Brandon D. Valentine" <bandix@looksharp.net> To: j mckitrick <jcm@FreeBSD-uk.eu.org> Cc: <freebsd-chat@freebsd.org> Subject: Re: Who writes the esoteric scientific Unix apps? Message-ID: <20011214135051.Q69086-100000@turtle.looksharp.net> In-Reply-To: <20011214170714.A13736@dogma.freebsd-uk.eu.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 14 Dec 2001, j mckitrick wrote: >We hear so often that high-end Unix worstations are used for advanced >scientific applications. The biggest of the big iron usually goes to >weather/finance forecasting and especially nuclear research. > >Who writes these apps? Specialized software companies or in-house >developers? Are they batch-based, command line apps or GUI driven? Here in the VU Center for Structural Biology[0] we use a mix of commerically developed and academically developed codes and applications. A great deal of the codes used in scientific computation are developed by the scientists themselves. It's always a treat when a grad student walks down to my office and asks to borrow a copy of K&R because she's been tasked with working on some computational problems. Scientists are not afraid of learning a little computer science on their way to finding solutions to their problems. One of the major codes we use for molecular dynamics simulation and structure refinement is a package called Amber[1] developed at UCSF by computational biochemists. This package sucks up most of our cycles here. It sells for peanuts to other academic institutions, and goes for big bucks if a commerical entity like a pharmaceutical company wants to use it. Amber is a batch-based app which we run on Beowulf clusters using Myrinet interconnects (low latency IPC is crucial to Amber performance and scaling) and big SGI NUMA boxes. We also make quite a bit of use of Insight, a commercially developed and marketed application sold by Accelrys[2]. This is an interactive GUI app which does visualization and also runs some structure calculation codes. It's an IRIX application and is mostly run on our O2s and Octanes. For visualization of complex biomolecules scientists here often use stereoscopic glasses which in combination with a toggle in the X app will display the molecules three dimensionally. We also make use of dino, moe, grasp, molmol, rasmol, curves, molscript, spdbv, msms, and numerous other packages to do NMR structure refinement and the like. The point being, there are every imaginable flavor of application in use, some open source, some closed, some commercial, some academic, some written in C/X/Motif, others written in f77, some are interactive, others are batch-based. The challenge to me and my colleagues is keeping this extremely heterogenous environment with one consistent, documented user interface. Most scientists don't have time to learn a different way of working every time they sit down in front of a new type of machine. They have enough to tackle just learning to use the new piece of anaylsis software. HTH. [0] - http://structbio.vanderbilt.edu [1] - http://www.amber.ucsf.edu/amber/ [2] - http://www.accelrys.com/ Brandon D. Valentine -- "Iam mens praetrepidans avet vagari." - G. Valerius Catullus, Carmina, XLVI To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011214135051.Q69086-100000>