On Friday 08 June 2001 09:38, SpeedMan wrote:
> On June 8, 2001 08:55 am, you wrote:
> > Question, in the machines mentioned below did you use the 80wire
> > connection cables on your hard drives or the old standard 40 wire cables?
> > From what I have read on the net this can become an issue with CDROM
> > drives and via chipsets if the 40wire cables are used.
>
> <SNIP>
>
> Excellent point Dennis. You are absolutely correct - standard IDE cables
> are a problem when used w/ VIA based Socket A motherboards.
>
> The systems I detailed earlier are my own personal equipment, and run high
> speed IDE cables on all controllers.
>
> My experience w/ SA/VIA boards goes beyond this though, as I own a small
> OEM and ship approx. 30 to 40 SA/VIA systems per month. We first started
> noticing problems late last summer - regardless of O/S - when running
> burners on the secondary master w/ standard cables. As soon as we switched
> to high speed IDE cables read/write errors disappeared.
>
> So, without a question I would recommend only using high speed IDE cables
> on SA/VIA based systems.
>
>
> Regards,
>
> SpeedMan
OK--do this
Put two disk drives on different channels on a 686B southbridge board
Activate DMA . Our kernel won't, so load kernel-linus2.4 or use another 2.4
distro or even win2K.
copy a partition of at least 100 Mb between them.
Try a diff. This error seems VERY reproducible. If it is reproducible here,
it makes one wonder about other problems. (I have experienced several
myself, and I used to swear by VIA).
http://www.au-ja.de/review-kt133a-1-en.html
But also, now a bug in the KT133A Northbridge has been confirmed.
Information on the bugs and the _fixes_ are at that site. The good news is
that there are fixes. The bad news is that I will have a difficult time
trusting again, and so will many others. Even those who have had clean
records may have just been dodging a hidden bullet.
http://www.neoseeker.com/forums/index.php?fn=view_thread&t=3394
Was VIA's official statement. The original testers did state the problem was
_exacerbated_ by the presence of a Creative Sound Card, but also were able to
show the bug without it.
http://www.au-ja.de/review-kt133a-2-en.html
They show a do-it-yourself patch for systems without creative soundcards.
They also say that the problem can occur with ANY PCI cards that generate DMA
traffic and may even occur without it.
Well I have watched Seagate ST34xxxA and ST38421A drives freeze solid on boot
with RH and Caldera and Mandrake 6.0 on a VIA MVP3. I have watched warm
reboots fail on FIC boards using KT133 with a nice Duron (I have two dead
KT133 boards in my collection, they simply stopped powering up though the
memory and all the cards and the processor were fine on other systems), so my
experience with VIA has been clouded with failures. The same could be said
for some others, like older PCChips boards, and the newer ASUS boards (when
did they start soldering in the BIOS--is someone new in ownership who decided
to cash in on the excellent ASUS reputation?). The fact is, you can't tell
much any more by brand name. You pays your money and you takes your chances.
When has it ever been different?
Probably VIA Chipsets will be better designed and less underpriced than in
the past, because of this huge scare. Still those who lost their work will
probably harbor bitterness for a long time, even though people repeat often
that Data Not Backed Up is Data Lost.
Civileme
