From gerson@vnet.IBM.COM Sat Jul 17 06:23:28 1993 Message-Id: <199307171425.AA28518@oscsunb.ccl.net> Date: Sat, 17 Jul 93 10:23:28 EDT From: gerson@vnet.IBM.COM To: chemistry@ccl.net, %chemistry@ccl.net Subject: SGI..HP..IBM Workstations From: Dennis J. Gerson, Ph.D. Consulting Scientist Engineering/Research and Quality Managment Practice --------------------------------------------------- Subject: SGI..HP..IBM Workstations Earlier in the long list of appends, Tom Pierce of Rohm and Haas reported that you need to "reallocate how memory is partitioned" on the IBM to improved performance for interactive tasks. Latter on, there were several appends to this bulletin board which discussed compiler options for improved performance For those trying to get their 6000's to perform "up to expectations" I would suggest you try three things before calling it quits- 1- Check into the documentation (the online INFO CD-ROM manuals) for tuning recommendations. 2- Ask you local IBM Representative to get you a copy of the International Technical Support Reports on Performance Tuning of AIX (these are red or blue covered books) 3- Call the 1-800# in the AIX documentation and ask the AIX/RISC System 6000 support line for assistance. In our group we are running UNIX on IBM, Intel-IBM, MAC and Sun Workstations and utilized SGI and HP systems attached to a mixed cluster of IBM/HP RISC boxes. We have "tweeked" AIX, SunOS and HP Unix kernel parameters to meet our needs using guidelines provided by the vendors. Regards, Dennis Gerson "the views expressed in this note are those of the author and not of the IBM Corporation or its subsidiaries" --------------------------------------------------------------- IBM Consulting Group |Tele:(214)280-1425 Fax:1486 1507 LBJ Freeway MS/160601 |IBMLINK: USIBM9YA @ IBMMAIL Dallas TX 75234 USA |Email: gerson@vnet.ibm.com From CLETNER@DESIRE.WRIGHT.EDU Sat Jul 17 14:02:02 1993 Date: Sat, 17 Jul 1993 19:02:02 -0500 (EST) From: CLETNER@DESIRE.WRIGHT.EDU Subject: Summary of Beginner References To: chemistry@ccl.net Message-Id: <01H0NR9QSO5E001FI6@DESIRE.WRIGHT.EDU> Hello Everyone, Earlier I posted a request for computational chemistry references for beginners. The response was great. Here is a summary of what I received. Three people sent extensive lists (the references they provided make up half this summary): Wilmon B. Chipman A.P. Heiner Peter Reinert The others that responded are: Shaun D. Black Eric G. Bradford Kimberley Cousins Greg Durst Dirk Huckriede Kestutis Sliupas David A. Winkler The most often mentioned book with 5 replies was: Clark, Tim, A Handbook of Computational Chemistry, Wiley, 1985. and in second with 4 replies was: Lipkowitz, K.B. & Boyd, D.B., Reviews in Computational Chemistry, Vol. 1,2,3,4, VCH Publishers 1990-1993. The remaining references were only listed once or twice. These are in alphabetical order: Advances in Biomolecular Simulations, AIP Conference Proceedings 239, 1991. Allen, M.P., Tildesley, D.J., Computer simulation of Liquids. Oxford Science Publ., 1987. Allinger, N.L., Yuh, Y.H., Lii, J.H., J. Am. Chem Soc., 111, 8551, 1989. Allinger, N.L., Kok, R.A., Imam, M.R., J. Computational Chem., 9:6 591, 1998. Boyd, D.B. & Lipkowitz, K.B. J. Chem. Educ. 59, 269-274, 1982. Brooks III, Charles L., Karplus, M., Pettitt, B.M., Proteins: A theoretical Perspective of dynamics, Structure and Thermodynamics, John Wiley and Sons, 1988. Burkert, Ulrich & Allinger, Norman, Molecular Mechanics, ASC Monograph 117, American Chemical Society, 1982. Cox, P.J. J. Chem. Educ. 59, 275-277, 1982. Crippen, G.M., Havel, T.F., Distance Geometry and Molecular Conformationn, Research Studies Press LTD, 1988. Engler, E.M., Andose, J.D., Schleyer, P.R., J. Am. Chem. Soc., 95:24, 8005, 1973. Karplus, M., Petsko, G.A., Molecular Dynamics Simulations in Biology. Nature, 347:631-637, 1990. Lesk, A.M., Protein Architecture, Oxford University Press, 1991. Magee, P., Probing Bioactive Designs, from an ACS meeting that may be of interest. McCammon, J.A., Harvey, S., Dynamics of Proteins and Nucleic Acids, Cambridge University Press, 1987. Rhodes, Gale, Crystallography Made Crystal Clear, Academic Press, 1993. Richards, W.G., Quantum Pharmacology, 1983. Rogers, Donald W., Computational Chemistry Using the PC, VCH Publishers, 1990. van Gunsteren, W.F., Berendsen, H.J.C., Computer Simulation of Molecular Dynamics: Methodology, Applications, and Perspectives in Chemistry, Angew, Chem. Int. Ed. 29:99201023, 1990. van Gunsteren, W.F., Mark, A.E., Eur.J. Biochem., 204 (1992), pp 947-961. von Kitzing, E., Methods in Enzymology, Modeling DNA Structures: Molecular Mechanics and Molecular Dynamics, vol.221, 1992. Weiner, S.J. et al, J. Computational Chem., 7, 230-252, 1986. Weiner, S.J. et al, J. Am. Chem. Soc., 106, 765-784, 1984. From lim@rani.chem.yale.edu Sat Jul 17 17:33:54 1993 From: Dongchul Lim Message-Id: <9307180133.AA28141@rani.chem.yale.edu> Subject: Re: SGI..HP..IBM Workstations To: chemistry@ccl.net (Computational Chemistry) Date: Sat, 17 Jul 93 21:33:54 EDT gerson@vnet.IBM.COM writes: > Earlier in the long list of appends, Tom Pierce of Rohm and Haas reported that > you need to "reallocate how memory is partitioned" on the IBM to improved > performance for interactive tasks. Latter on, there were several appends > to this bulletin board which discussed compiler options for improved > performance > The following may not belong to this thread, but it may be helpful to some readers. There was a review on IBM RS/6000 system (SunWorld Page 53-58, February 1993 issue). RISC workstations run AIX, IBM's coporate UNIX implementation. The AIX has some innovative features, such as dynamically linked libraries at the kernel level and the pagable kernel. This indicates that RISC is a good choice for a real time work. However, the AIX is not a UNIX which we got used to. Especially, its virtual memory manager (VMM) is very different from other UNIX systems. For example, when the program requests an allocation of memory (using malloc), the AIX returns a pointer to the memory location like other UNIX system. But unlike ordinary UNIX, the memory is not actually allocated in the AIX until the program accesses it. If the memory is not available at that time, the AIX generates a SIGDANGER signal and finally kills the program to free the space. What's the difference? In other UNIX systems, programmers know if the memory is available when they try to allocate it. On AIX, they don't until they get a SIGDANGER signal from the operating system. Other than memory menaging problem, the review was mostly praising RISC's great performance. -DCL