Date: Fri, 27 Oct 1995 01:19:56 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@GndRsh.aac.dev.com> To: davidg@Root.COM Cc: hardware@freebsd.org Subject: Re: Asus PCI/I-P55SP4 motherboard Message-ID: <199510270819.BAA05026@GndRsh.aac.dev.com> In-Reply-To: <199510262308.QAA00392@corbin.Root.COM> from "David Greenman" at Oct 26, 95 04:08:22 pm
next in thread | previous in thread | raw e-mail | index | archive | help
>
> >> I'd like to hear from anyone with any experiance with this motherboard. I'm
> >> considering using it in wcarchive, and I'd like to know about performance and
> >> reliability.
> >
> >It has run make world's with a 2940, a few card compatibility tests etc,
> >looks to be okay. It is benchmarking slower than the PCI/I-P55TP4XE, but
> >I suspect that is because the book says it will run the 100Mhz chip with
> >70nS memory, which means they stuffed a wait state in it. I need to get
> >the Sis data books so I can tweak the BIOS values and know exactly what
> >it is actually doing as there are about 12 memory timing parameters.
>
> Can you be more precise about how much slower it is? Are we talking 2-3%,
> or are we talking 10-15+% slower?
You talking a wait state, so 10% is about what I have seen so far. I have
a very small amount of data at this time, but that is what it says.
> >Oh, and that board supports up to 200MHz chips per the 1.2 Errata sheet
> >stuck in my manual, though ASUS is not stuffing the VRM socket in the
> >board which means you can only go to 150Mhz (All chips faster than 150Mhz
> >use a 2.9V VRM power source).
^^^^
My error here, VRM is 2.5V.
>
> Hmmm...their Web page says it supports 167Mhz...
ASUS web pages are created way before real product happens, and they never
seem to update them. My data on ASUS is 2x to 4x more accurate than the
web site, as I deal with _real_ product day in day out.
Another part of the errata I didn't notice when repling to you but noted
later was ``the VRM connector is an optional item at this time as VRM's
are expensive and not expected to be avaliable until 2H 1996''. I take
this with a grain of salt and add the comment ``P6 starts shipping
Q1/Q2 1996, sounds like VRM P54 chips are going to be another P24T story
and something we should not count on.
I will also add a hard cold fact, ALL Pentium chips with the core running
above 100Mhz are _seriously_ restricted by the cache to cpu bandwidth,
so unless you have things that run out of the internal cache for reasonable
lengths of times these chips are a waste of time. Pipeline Burst cache
modules help a lot, infact a 100Mhz chip with a PB cache will run a make
world faster than a 133Mhz chip with an async cache.
If you have some specific tests you would like to run on the ASUS PCI/I-P55SP4
board please feel free to contact me.
--
Rod Grimes rgrimes@gndrsh.aac.dev.com
Accurate Automation Company Reliable computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199510270819.BAA05026>