Xref: utzoo alt.folklore.computers:7677 comp.unix.internals:1218 comp.misc:10730 Path: utzoo!utgpu!cs.utexas.edu!uunet!mcsun!ukc!dcl-cs!aber-cs!athene!pcg From: pcg@cs.aber.ac.uk (Piercarlo Grandi) Newsgroups: alt.folklore.computers,comp.unix.internals,comp.misc Subject: Re: Jargon file v2.1.5 28 NOV 1990 -- part 5 of 6 Message-ID: Date: 4 Dec 90 16:07:37 GMT References: <1YbxGQ#2fbT353y6xKD8DT83C4bFDpV=eric@snark.thyrsus.com> <1990Nov30.172512.5282@sctc.com> Sender: pcg@aber-cs.UUCP Organization: Coleg Prifysgol Cymru Lines: 46 Nntp-Posting-Host: odin In-reply-to: pst@ir.Stanford.EDU's message of 1 Dec 90 21:14:40 GMT On 1 Dec 90 21:14:40 GMT, pst@ir.Stanford.EDU (Paul Traina) said: pst> In article peter@ficc.ferranti.com pst> (Peter da Silva) writes: pst> Actually, those of us who still thing REAL UNIX means Version 6 or pst> Version 7 find this particular comparison wonderfully appropriate. pst> Version 7 UNIX at Berekeley supported 60 users on an 11/70. [ ... ] pst> The 386 box on my desk at work is a comparable machine, with 5 pst> times the RAM of the old 11/70, but more than 10 users kill it pst> dead. Note that the number of users that an 11/70 could support depended critically on whether they were using ed or vi (line or character at a time IO), and on whether you had fast or slow swapping. Swapping is good -- most of the time timeharing users are thinking (say one user in 10 is usually "active"). If swapping in takes less than half a second it is imperceptible. Unfortunately designing a good swapper is a tough thing apparently, and the UNIX swapper has not been redesigned for VM; it is still the one used to swap 64k images. pst> You're comparing CPU performance to I/O performance. [ ... ] Back pst> when there were REAL(tm) computers like 780, a lot of time and pst> energy went into designing efficient I/O from the CPU bus to the pst> electrons going to the disk or tty. [ ... ] Sure OS's and apps have pst> gotten bloated, but when you put a chip like the MIPS R3000 on a pst> machine barely more advanced than an IBM-AT you end up with a toy pst> that can think fast but can't do anything. No, no, no, no, no, no, no. The IO bandwidth of a typical 386 is equivalent or better than that of any UNIBUS based machine, and, in practical terms, equivalent to that of MASSBUS based ones. You can get observable raw disc data rates of 600-900KB/s and observable filesystem bandwidths of 300-500KB/s under SVR3.2 (with suitable controllers and a FFS of some sort). This is way better than a PDP-11. There are many reasons for which UNIX has become even more obscenely inefficient; mostly it is just plain old lack of hard thinking (read "lack of design talent"). I love to repeat my old line: talented (japanese) process engineers are easier to find than (american) OS designers, so there is abundant supply of high density RAMs and high waste OSes. -- Piercarlo Grandi | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk Dept of CS, UCW Aberystwyth | UUCP: ...!mcsun!ukc!aber-cs!pcg Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.uk Brought to you by Super Global Mega Corp .com