I've only ever updated the BIOS on two of my older machines.....and those were well over a decade ago.
My ancient, 2002 Dell Inspiron 1100 needed it because under XP, the 'Brookedale'-cored, 82845-G chipset (which included the 'Extreme' graphics adapter) permitted a maximum of just 1 MB of VRAM. It seemed to run fine for XP, but no Linux distro of the time would play ball with that.....so the update I applied allowed the user to boost that all the way up to a heady 8 MB of VRAM!
By visiting a site called Bay-Wolf.com, you used to be able to perform this upgrade via a downloaded, teeny ISO image. This merely contained the Dell-supplied .exe file, a batch file and a 'ReadMe', though turning it into an ISO made it 10x simpler to use .Burn it to a CD, boot from that CD, follow the instructions.....30 seconds later, it was 'done & dusted'. It was the easiest & most foolproof way of doing said update I'd ever found, given that everyone else at that time expected you to mess about with floppies within a DOS environment.
This upgraded me from A06 all the way up to A32.....and also fixed a host of other niggling little issues, most of which I'd grown so used to I no longer even noticed them.
God above knows why, but i still have this ISO file kicking around in my archives.....despite the machine in question being sent to the great dumpster in the sky several years ago..!
This site is long since defunct. Visiting the URL now will merely bring up the "This site cannot be reached" message...
~~~~~~~~~~~~~~~~~~~~~~~~~
The other one was the elderly Compaq Presario desktop rig I "inherited" from my sister around 2012 when she upgraded to Win 7. This was a 2005 machine, built using the very last supplies of genuine Compaq parts immediately after the sale to HP......Compaq components, although an HP-spec'd mobo from MSI.
It came with a single-core,
AMD Athlon64 3200+ on
Socket 939. I soon discovered it was possible to turn it into a dual-core machine, since the same
ATI RS482 chipset not only supported the
Athlon64 FX 'enthusiast' CPUs, but also the early
Athlon64 X2 dual-cores. In late 2012, I snagged a first-gen X2 3800+ dual-core off eBay for the princely sum of GBP £7.23, in excellent condition!
This would drop straight into the same CPU Socket 939.....but required a BIOS update to utilize the dual-core's architectural advantages.
~~~~~~~~~~~~~~~~~~~~~~~
With help from an acquaintance, we performed this particular upgrade from within 'Puppy', making use of the Linux
flashrom utility, along with plenty of reference to the Arch wiki.
The one fly in the ointment was the mobo itself. This was designated MS-7184 by MSI, but you could NOT find it in their listings. It was one of those boards made under contract to an OEM; so far as MSI were concerned, once made & delivered to HP they were nothing to do with them any more, and they washed their hands of them, passing all responsibility to the customer.
However, the only BIOS upgrade HP offered was the very one I already had! So, we hatched a plan...
My acquaintance on the Puppy forums did some further digging.....and discovered an almost-identical MSI board designated MS-7093. In every respect, the specs were identical.....apart from this one offering 4 SATA1 ports, where the MS-7184 offered just two. (This was built around the time of the transition from IDE connectors to SATA ones, so the mobo in fact supported both). Turned out the RS482 supported up to 4 anyway, but HP had 'downgraded' the firmware or something like that. So.....
.....we decided to take the risk, since the BIOS upgrade for the MS-7093 included the required support for the X2s.
~~~~~~~~~~~~~~~~~~~~~~~
All went very smoothly using 'flashrom'. I had an interesting half-hour immediately after re-booting; the fans were screaming away at full blast, and I couldn't get into the BIOS.....whoever had uploaded the BIOS file had locked it with a password! I spent the next 20 minutes feverishly searching various websites for any possible passwords which would let me unlock the darned thing.
We eventually found one, and after entering the BIOS found the thing appeared to have been zipped up with some kind of weird custom profile. I spent nearly 10 minutes resetting everything to my own preferences.....and all was finally serene once more, and ticking-over like a well-oiled watch.
Yes, you DO take risks with this kind of operation, though in my own case we were having to do a bit of a 'Frankenstein' job anyway, simply to get things functioning in the desired manner with what was available. But it DID work, and continued to behave itself perfectly until the mobo eventually expired from dried-out caps nearly 7 years later.
Ever afterwards, any system queries always reported an MS-7093 mobo, despite that it was, in fact, a "non-existent" MS-7184..!
Mike.
