Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
I was receiving the same message and updated to the 1013 BIOS. The notes
indicated that it dealt with erroneous messages to this effect. I've not
been warned by Betty since.
"Paul" <nospam@needed.com> wrote in message
news:nospam-1204052301100001@192.168.1.178...
> In article <1113339417.288885.262340@l41g2000cwc.googlegroups.com>,
> aberger@u.washington.edu wrote:
>
>> Paul wrote:
>> > In article <O9ydnTtGKNJeR8TfRVnygA@giganews.com>, "Rudy Kazuti"
>> > <rudykube@comcast.net> wrote:
>> >
>> > > Every once in a while on boot up I get the message 'system failed
>> due to cpu
>> > > overclocking'. I'm not overclocked, Everything is running at auto.
>> CPU speed
>> > > is right, 200. Multiplier is correct at 11. All voltages are
>> normal. After
>> > > the message it boots right into Win 2000 Pro and everything is
>> fine.
>> > > AMD 3200+ /400 fsb
>> > > PC3200/400 ram 512
>> > > 120gig WD IDE Hdd
>> > >
>> > > Rudy
>> >
>> > I bet you are not "Prime95 clean". Get a copy of Prime95 from
>> > mersenne.org and run the "torture test" option. I am willing to
>> > bet you get rounding errors after 30 minutes or so.
>> >
>> > On my system, the "system failed..." error messages stopped,
>> > after I fixed my memory. You don't say it in your posting, but
>> > I bet you are using CAS3 memory. I cured mine by buying some
>> > Crucial Ballistix CAS2 memory, but any other PC3200 CAS2
>> > memory is likely to work the same.
>> >
>> > Another option, is to use a hacked "Command Rate 2T" BIOS
>> > (otherwise know as Command Per Clock Disabled or CPC Off).
>> > You can search over on nforcershq.com, for comments on CPC.
>> >
>> > If you use a CPC Off BIOS, you will find operation at
>> > FSB400/DDR400 becomes stable, but if you benchmark with
>> > Sandra, your memory bandwidth will have dropped a bit as
>> > a result.
>> >
>> > The hacked BIOS is a "zero cost" option. I have experimented
>> > with the "Trat" hacked BIOS, and had no problems flashing that
>> > BIOS into the machine. If you pick up a hacked BIOS, make
>> > sure to read some postings from people who have tested it
>> > first. Purchasing a BIOS Savior from ioss.com.tw is
>> > reasonably cheap insurance ($25 or so) against a flash
>> > going bad.
>> >
>> > If you choose to ignore the problem, you might occasionally
>> > get a memory error, in which case the worst result could
>> > be a corrupted registry. If you continue to run your system
>> > in its current condition, make sure you are rigorous about
>> > backing up your data.
>> >
>> > HTH,
>> > Paul
>>
>> Why can't he go into the BIOS and just set the CAS timing to 3 wait
>> states instead of 2?
>>
>> ab
>
> I tried relaxed DRAM timing and it didn't help at all. The fact
> that other people have noticed changing Command Rate to 2T
> (not the same thing as CAS) makes it work better, tells you
> that perhaps there is long prop delay versus load on the
> Nforce2 address busses. Since Asus didn't see fit to put
> the Command Rate setting as a BIOS option, it takes flashing
> a hacked version of the BIOS in order to set Command Rate to
> 2T (= CPC off). The Asus BIOS are all set up for
> Command Rate 1T (= CPC on), so there is no extra cycle of
> setup time on the address bus.
>
> I have read at least one report of some CAS3 memory that
> worked, plus a lot of complaints from people with symptoms
> like mine. If CAS3 memory will work for you, then great.
> If not, either a "CPC off" hacked BIOS, or the purchase
> of some CAS2 DIMMs, might just help.
>
> It is possible that some memory which is branded as CAS3
> memory, but which is made with CAS2 capable memory chips
> on it, may be the right ingredient to a working Nforce2
> solution. But you would have to work in a computer store,
> to get enough info about what is "underneath the heat
> spreader", to be making informed purchases like that.
>
> Paul