Subject: Re: ballpark ratio users:cpu for a [345]86 NetBSD box?
To: None <perry@imsi.com>
From: Dave McGuire <mcguire@rocinante.digex.net>
List: current-users
Date: 09/27/1994 15:36:59
On September 27, you wrote:
> Pardon, but a Pentium with a PCI bus has far better context switching,
> far better IO, and far better speed. What is left that the Sun 4/470,
> which was designed as a server, not as a multiuser machine (which is
> why it has too few MMU contexts) can beat the Pentium machine on?

  I'd argue that.  I will freely admit to having never *tried* to put
60 people on a PC, because the idea is just plain silly in my opinion.
The 4/400 CPU has 64 contexts, the largest number in the sun4 series
excluding the 4/600 (different animal altogether) and as such does
indeed make a wonderful multiuser machine.  That I *have* tried.
Again, I'm not trying to start a flame war here, so please don't turn
it into one.

> I can also pick one of those up in a reasonable configuration for
> $2,000.

  Really?  Now *that* comes as a surprise to me.  If they're down that
low these days, I think I might have to look into it.  That'd be quite
cool.

> Its not quite the same O/S or application, but I'll note that
> Compuserve has been successfully experimenting with hundreds of users
> on single PCs running BSDI.

  This I would like to see.

> The main problem I could see with a NetBSD Pentium box right now is
> that NetBSD won't max out the disk because the clustering stuff isn't
> in yet. However, the hardware is as capable as anything you can
> find. The Pentium is an UGLY chip in terms of instruction set, but the
> notion that the machine can't compete is misplaced. Try a benchmark
> before claiming otherwise.

  I will agree that it has come a long way, mostly hampered by having
to be crippled enough to still run DOS.  If it weren't for that, I
think they'd be running rings around bigger systems.  No, I haven't
tried a benchmark.  But it's a matter of architecture.  PCs just
simply weren't designed for it.  Recently there was a conversation in
comp.sys.dec where some fellow happened upon a MicroVAX III.  In
asking what it was, some guy explained that it offered about the same
performance as a 386DX/25.  Well, I've run both at work, and I have
both at home.  I can put 25 people on a MicroVAX III if it's properly
maintained.  I'd like to see someone try it with a 386DX/25.

  Out of curiosity...I honestly don't know much about the Pentium's
instruction set, beyond the obvious 8086isms.  What is ugly about it?


                           Regards,
                            -Dave McGuire
                             Operations
                             Digital Express Group, Incorporated
                             mcguire@digex.net