biosoft packages: performances on different boxes?

Stefano Ottani stefano at frodo.ciam.unibo.it
Wed Dec 13 13:22:58 EST 1995


	We are about to start some new projects on the modeling of natural
and synthetic macromolecular systems. We are planning acquisition of a new
machine to run molecular mechanics and dynamics packages and also
crystallographic packages for solution and refinement of protein structures
(viz. GROMOS, MacroModel, AMBER, X-PLOR, the CCP4 suite...). 
	After consideration of economic factors and performances, we ended up
with three possible machines all equipped with 128Mb RAM,. 2Gb HD and
3D/Z-buffer level graphic boards: 

    a) IBM RISC/6000 390H 1Mb Cache L2 - 260 specfp92 
    b) DEC Alpha station 300/266 2Mb cache - ~400 specfp92 
    c) HP 9000 C110 (PA-RISC 7200/120mhz) - 269 specfp92

	For several reasons we could not consider another obvious candidate
like the SG Indigo 2 :-(. 
	From the above reported figures DEC Alpha seems the best choice. 
        However we are quite concerned about the meaning of such figures in
real life. 
	We would really appreciate any information regarding performances of
these boxes with the above mentioned or analogous software.
	Thanks in advance for any helpful suggestion

                           Stefano Ottani
 ____________________________________________________________________________
|                                    |                                       |
|  Stefano Ottani                    |  E-mail: stefano at frodo.ciam.unibo.it  |
|  Centro Studi Fisica Macromolecole |                                       |
|  c/o Dip. Chimico "G.Ciamician"    |  Phone: +39-51-259476                 |
|  Via Selmi, 2                      |                                       |
|  I - 40126     Bologna (Italy)     |  Fax:   +39-51-259456                 |
|____________________________________|_______________________________________|






More information about the Bio-soft mailing list