home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!rde!ksmith!keith
- From: keith@ksmith.uucp (Keith Smith)
- Subject: Re: IBM AS/400 is the world's slowest computer
- Organization: Keith's Computer, Hope Mills, NC
- Date: Tue, 22 Dec 92 20:21:15 GMT
- Message-ID: <1992Dec22.202115.15118@ksmith.uucp>
- Followup-To: comp.misc
- References: <1992Dec18.165950.15013@webo.dg.com> <1992Dec20.221103.13995@ksmith.uucp> <1992Dec21.220736.27684@rchland.ibm.com>
- Lines: 51
-
- Note follow-up line.
-
- In article <1992Dec21.220736.27684@rchland.ibm.com> kunkel@vnet.ibm.com writes:
- >The B20 is the second smallest AS/400 ever shipped and was announced
-
- So What
-
- >4.5 years ago. The current largest model is about 20 times faster
- ^^^^^^^^^^^^^^^
- So do you wish to compare the most expensive AS/400 you can buy to a
- 486/33. Give me a break, or better yet give me some compile, & sort
- times for say 1000 line RPG programs and 50-100MB database files using
- the native compiler and #GSORT. Even better yet I can mail you the one
- I used & We'll compare
-
- >(TPC-C type benchmark). The B20 is also a multi-user system. What
-
- So was the SCO Unix on a 486/33 I tested it against. The SCO box also
- had *MORE* Crap hung off it.
-
- >were the other 10 or 20 users doing when you compiled? Any batch
- >jobs running? TPC-C is a system benchmark. A compile is a CPU
- >benchmark. There is a different. The AS/400 system has a couple
-
- Yep, Transactions test the disk & I/O more than the CPU. I shant argue
- with you there. However, I did the comparison *AFTER* We were switched
- over when absolutely NOTHING was going on on the AS/400. There were
- only 3 Terminals and 1 Printer Varied on at the time. Only one Terminal
- was active. On the '33 at the same time there were 2 Highspeed modems
- doing a file transfer, and 3 or 4 people doing Word Processing. Nothing
- Heavy there either.
-
- >(many hand fulls on the high end) of cheap, off-the-shelf micros
- >thrown in as a cost effect way of off loading I/O processing from
- >the CPU.
-
- So does my 486/33 (which was upped to a 66 last month or so). I've got
- a Digiboard C/X system with 4 286 CPU's in it (You know, the stuff your
- brothers hang on the RS/6000). The new ones use a RISC I think. The
- Disk Controller has a 16Mhz 68000 Chip. And all the Drives Are SCSI (Oh
- yea, that needs cpu's huh). That's 5 more. Oh and of course the SCSI
- tape (8mm) has one too. That makes a total of 11 dedicated cheap off
- the shelf I/O processors, and you ain't just whistling dixie about
- offloading. I did a simple loop test on the Digiboard vs a pair of
- 16550's on the bus (ISA SIO card), and the System load difference is
- amazing. Like 1/10th. The SCSI card takes care of the mirroring
- transparently too.
- --
- Keith Smith uunet!ksmith!keith 5719 Archer Rd.
- Digital Designs BBS 1-919-423-4216 Hope Mills, NC 28348-2201
- Somewhere in the Styx of North Carolina ...
-