friedell%chimera@HARVARD.HARVARD.EDU (Professor Mark Friedell) (11/22/85)
I've had a 32C for almost 2 years and I have a love/hate relationship with this machine. For compute-bound jobs, the 32C is remarkably fast, but the operating system is almost unbelievably poor at switching processes -- I'm told that the problem lies in the memory-management scheme, but I don't know for sure. For developing software, the 32C could be wrong choice: the user interface is just responsive enough for one (really, one!) user. Two users get very angry. With three users you might just have to find some other time to use the machine. My 32C is on an ethernet with lots of other machines, including two Vax 11/780s and a bunch of Suns. The 32C is faster than the 11/780 for many compute-bound jobs (like the image-rendering work I do), but, considerably less responsive than the Suns when two or more users ATTEMPT to run interactive programs (like text editors). Mark Friedell Center for Research in Computing Technology Harvard University (617)495-5841
info-ridge@ucbvax.UUCP (11/23/85)
We are considering buying a couple of Ridge 32 systems, to be used for general software development, but the project is contending with a solution consisting of a network of Sun workstations (which appear to be more 'mature', at least as far as software is concerned). We don't know much about the Ridge and about ROS since they both have been introduced in France only recently, so I would like to ask anybody willing to reply to send me his/her personal critical evaluation of Ridge 32/ROS (hardware, software, reliability, interactivity, overall performance,...). Please consider that any data you will volunteer *will be used* in the decision phase of our project, and will be read by *several* people. Hoping to receive many answers, I would suggest that you mail these to me personally; I'll post a digest on the list when I have received enough stuff. (is this ok with the moderator ?) Many thanks in advance for those who will reply. Didier Giralt (Toulouse, France) ...!mcvax!inria!lasso!didier
info-ridge@ucbvax.UUCP (12/07/85)
We have a Ridge here and it's been very disapointing. If you can get sun's your better off. I can give you more specifics but the people here other than myself support it and can give you lots of the ugly details. -------------------------------------------- Pete Delaney - Rocky Mnt. Unix Consultant Phone: (49) 89 92699-139 European Computer-Industry Research Center UUCP: mcvax!unido!ecrcvax!pete ArabellaStrasse 17 UUCP Domain: pete@ecrcvax.UUCP D-8000 Muenchen 81, West Germany X25: (262)-45890040262 CSNET:pete%ecrcvax.UUCP@Germany.CSNET Login: <to be provided?> ------------------------------------------------------------------------------
info-ridge@ucbvax.UUCP (12/10/85)
In article <ecrcvax.8512052249.AA09223> you write: >We have a Ridge here and it's been very disapointing. If you can get sun's >your better off. I can give you more specifics but the people here other >than myself support it and can give you lots of the ugly details. > > >-------------------------------------------- >Pete Delaney - Rocky Mnt. Unix Consultant Phone: (49) 89 92699-139 >European Computer-Industry Research Center UUCP: mcvax!unido!ecrcvax!pete >ArabellaStrasse 17 UUCP Domain: pete@ecrcvax.UUCP >D-8000 Muenchen 81, West Germany X25: (262)-45890040262 >CSNET:pete%ecrcvax.UUCP@Germany.CSNET Login: <to be provided?> >------------------------------------------------------------------------------ I fear the powers-that-be won't like the user community to hear such things. Would you mind sending a cancel message. I would appreciate receiving a positive ack that you have canceled this "offending" article.
info-ridge@ucbvax.UUCP (12/11/85)
The Ridge was purchased here as a number crunching computer to offload large
floating point background jobs from our Vax 750. For this application I would
say that the Ridge has been a success. There have been bugs found related to
numerical computation on the Ridge but 4.2 BSD's f77 compiler for the VAX has
hardly been an example to emulate in that respect.
However for interactive use the Ridge is a disaster. At load average 3 the
interactive reponse of the Ridge feels slower than the Vax does at load
average 10. This is depressing for a machine that is supposed to be a fair
amount faster than the Vax 750. The Ridge has 7 Mb of memory on it so I don't
see how paging could be the problem.
Our next system is a Sun 3 that will be dedicated to image processing. We'll
see how that compares. The 68020 is reputed to be a hot chip. Sun now will
give source to (most) of their software to educational institutions for $1000.
As I have an undying hatred of binary-only software distributions this gives me
another reason to prefer the Sun. The Microvax II was the main contender.
Source for Microvax Ultrix for an educational institution is now $750. However
we went with the Sun 3 on the assumption that it was faster and the display
software somewhat more mature.
Bill Sebok Princeton University, Astrophysics
{allegra,akgua,cbosgd,decvax,ihnp4,noao,philabs,princeton,vax135}!astrovax!wls