Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
"Paul" <nospam@needed.com> wrote in message
news:nospam-1806040907520001@192.168.1.177...
| In article <2jfkk3F103truU1@uni-berlin.de>, "Kylesb"
<me@privacy.net> wrote:
|
| > "Paul" <nospam@needed.com> wrote in message
| > news:nospam-1706042329030001@192.168.1.177...
| > | In article <vJsAc.70$7d2.3@clgrps13>, "peter" <peter@telus.net>
| > wrote:
| > |
| > | > I have never heard of the "BIOS save death" bug
| > | > what is it??
| > | > peter
| > |
| > | There are 75 hits here:
| > |
| >
http://groups.google.com/groups?q=%22bios+save+death%22&ie=ISO-8859-1&hl=en
| > |
| > | Try the three terms BIOS save death and click "search for all
| > terms".
| > |
http://nforcershq.com/forum/search.php
| > |
| >
http://nforcershq.com/forum/viewtopic.php?t=37481&highlight=bios+save+death
| > |
| > | On most motherboards, CMOS settings are stored in a small RAM in
the
| > | Southbridge. My personal theory, is the small RAM in the Nvidia
| > | Southbridge chip is non-functional, so the BIOS writers decided
to
| > | store the CMOS settings inside the BIOS chip instead. A failure
| > | of an attempt to write to the BIOS chip (when you do Save and
Exit),
| > | followed by the BIOS doing a hardware reset, results in a
corrupted
| > | chunk of settings in the BIOS chip. Now, I have nothing to back
| > | this up, other than to say that it is highly unlikely that if
| > | the RAM in the Southbridge was working properly, there is no
good
| > | reason for the BIOS chip itself to be corrupted, unless it is
being
| > | used for a workaround as I outlined.
| > |
| > | HTH,
| > | Paul
| > |
| >
| >
| > If this were true, then removing the battery (or jumpering the
"clear
| > CMOS" jumper) would not clear CMOS settings, keep in mind data
written
| > to the BIOS chip is permanent (until overwritten). I'm pretty
certain
| > the CMOS settings are stored in ram, maybe that ram is found in
the
| > Southbridge chip, I am not well versed on the NF2 chipset designs
to
| > be certain. I've read of the BIOS save death bug, I think it may
be
| > related to unstable BIOS settings or PS problems that create bad
JUJU
| > on the mobo during CMOS write operations, just my SWAG on the
subject.
| > The SIS735 chipset has a "lost CMOS" problem, which does not
result in
| > BIOS death, but rather, a resetting of the CMOS settings, and some
| > have found the problem exists due to the temperature of the
chipset at
| > power on.
|
| But the people affected by the bug, cannot recover by removing the
| battery or by following the "reset the CMOS" procedure. I can only
| conclude from this, that "Save and Exit" in the BIOS is doing
| something else. There is apparently an observable delay during
| this time, and it doesn't take time to write to the CMOS. Flashing
| a BIOS takes time, because writing an EEPROM has a fair delay per
| location, at least compared to the CMOS RAM in the Southbridge.
|
| I've never seen a credible explanation of the problem, so we're
| left to guess. Maybe someone who owns both an A7N8X Deluxe and
| an A7N8X-E Deluxe could comment on whether the delay after
| doing a "Save and Exit" is the same on both boards or not.
|
| Since every bit of info from Nvidia would be covered by an
| NDA, we're not likely to get an explanation or an admission
| of guilt.
|
| This article, as far as I'm aware, doesn't appear to be true.
| Setting the FSB to 100 via jumper, or a repaired BIOS from
| Nvidia didn't fix the problem either. And it is not strictly
| an overclocking issue either, as for some people it dies at
| stock settings. The "workarounds" trumpeted by some people, don't
| work for all, so product is still being RMAed to this day.
|
|
http://www.theinquirer.net/?article=8250
|
| Paul
Yes, I too have noticed an extended delay during CMOS save operations
at times, typically when changing ram timings and/or FSB settings,
critical settings, as if some testing was underway. It would actually
be fairly simple to test your theory, one would have to boot w/o any
memory manager, use debug to read the entire BIOS into a file, reboot,
make some CMOS changes, save the changes, then read the BIOS into
another file and then do an FC on the files.
--
Best regards,
Kyle
Facebook
Twitter
Instagram