fkittred@bbn.com (Fletcher Kittredge) (03/29/91)
In article <91086.114156X03@psuvm.psu.edu> X03@psuvm.psu.edu writes: >HP 9000/700 was supposed to have been announced officially yesterday. Can >someone who knows this post some information: performance, price, software >availability,... Software availability is not an issue, since HP 9000 S800 binaries "should" run unmodified on HP9000 S700. Unlike the IBM 6000 the Series 700 is compatable with the previous revision of the architecture, so there is a large body of software already available to run on the Series 700. Note that you can run Series 800 binaries on a Series 700, you can not run Series 700 biniaries on a Series 800 (unless you cross-compile). We took our large, sophisticated application over to the S700, and it ran like a charm. We took our large, complex development environment, which includes many custom tools, over to the the HP 9000 S700, and it ran many times faster. nice box... regards, fletcher Fletcher Kittredge Platforms and Tools Group, BBN Software Products 10 Fawcett Street, Cambridge, MA. 02138 617-873-3465 / fkittred@bbn.com / fkittred@das.harvard.edu
perry@hpfcdc.HP.COM (Perry Scott) (04/03/91)
>We took our large, sophisticated application over to the S700, and it ran >like a charm. We took our large, complex development environment, which >includes many custom tools, over to the the HP 9000 S700, and it ran many >times faster. The PA-RISC 1.1 compiler has a few tricks for getting more performance. If you have the sources, you might try recompiling. It may run a little faster, if that matters to you. But, as you say, the 700 runs 800 binaries just fine. In fact, most of the commands (/bin, /usr/bin) are 800 binaries. Do a "file /bin/*" on a demo machine and see what I mean. I have not yet met an 800 executable that won't run on a 700. >nice box... Thanks. >regards, > >Fletcher Kittredge Perry Scott HP Ft Collins
jbc@hpcupt3.cup.hp.com (Jeff Caldwell) (04/04/91)
> Software availability is not an issue, since HP 9000 S800 binaries "should" > run unmodified on HP9000 S700. Unlike the IBM 6000 the Series 700 is > compatable with the previous revision of the architecture, so there is a > large body of software already available to run on the Series 700. Note > that you can run Series 800 binaries on a Series 700, you can not run > Series 700 biniaries on a Series 800 (unless you cross-compile). Yep. HP's talking REAL compatibility here. We're a company that makes sure our current customers are taken care of. Of course, with these new machines, we don't mind bringing a few more customers from under the sun :-) > We took our large, sophisticated application over to the S700, and it ran > like a charm. We took our large, complex development environment, which > includes many custom tools, over to the the HP 9000 S700, and it ran many > times faster. Glad to hear it. I'm sure recompiling these with the 700 compilers will show even a greater gain. > nice box... Thanks. -Jeff Caldwell | HP California Language Lab
milburn@me10.lbl.gov (John Milburn) (04/04/91)
In article <5570592@hpfcdc.HP.COM> perry@hpfcdc.HP.COM (Perry Scott) writes: >But, as you say, the 700 runs 800 binaries just fine. In fact, most of >the commands (/bin, /usr/bin) are 800 binaries. Do a "file /bin/*" on a >demo machine and see what I mean. I have not yet met an 800 executable >that won't run on a 700. I don't doubt what you are saying, but based on history, I think your evidence is flimsey. This is from my 400t, running hp/ux 7.0: me10:milburn 40 %) file /bin/ls /bin/ls: s200 demand-load executable -version 2 Well, at least I know it's a Motorola :-) -jem -- John Milburn milburn@me10.lbl.gov (415) 486-6969 ..And since the stench of death will always attract flies and vermin, the arrival of Geraldo was perhaps inevitable.
perry@hpfcdc.HP.COM (Perry Scott) (04/09/91)
> >>But, as you say, the 700 runs 800 binaries just fine. In fact, most of >>the commands (/bin, /usr/bin) are 800 binaries. Do a "file /bin/*" on a >>demo machine and see what I mean. I have not yet met an 800 executable >>that won't run on a 700. > >I don't doubt what you are saying, but based on history, I think your >evidence is flimsey. OK, it's flimsey. Try this: # file /usr/sam/bin/get_disks /bin/ls /usr/sam/bin/get_disks: PA-RISC1.1 shared executable /bin/ls: s800 shared executable The core commands group, which sits across the aisle from me, made a concious decision not to to Snakeify the commands. The ROI just wasn't there - two sets of binaries, incompatible binaries across NFS mounts, etc weren't worth the speed advantage. So, they took the kernel team at its word, and shipped the 8.0 800 command set. The test suite actually reports fewer problems on the 700 than the 800, so everything appears to be OK on the compatibility front. >Well, at least I know it's a Motorola :-) Yeah, point taken. Have you found a 200 that won't run binaries produced by today's compilers ? Looks to me like the 200->300->400 story applies to 800->700. HP supports old iron. >-jem Perry