NCR SDMS BIOS upgrade on Plato motherboard
: I get nonsense like "We have no plans on releasing a new version..."
: blah blah blah. I told them I'm not interested in what they release,
: I just want to patch MY motherboard. I'm not holding my breath for a
: response.
: I'm REALLY getting fed up with this runaround I'm getting between
: Symbios and Intel. Symbios's position seems a bit more reasonable
: (after all, they can't control what motherboard vendors do). If Intel
: and Symbios/NCR had a falling out, I don't want to be the one getting
: the shaft.
Well, Symbios (AT&T, NCR, whatever they're calling themselves right now)
can't write flash update programs for every motherboard. That's silly.
However, they do make the BIOS available freely for download. If Intel
didn't have a braindead update program or braindead policy which requires
them to encrypt the BIOS, this wouldn't be a problem. You could just
patch in the SDMS BIOS into the Intel update. But, noooo. And, Intel
won't tell anyone how the BIOS updates are encrypted, that would defeat
the purpose of encryption! Duh.
: I understand Asus makes some fine motherboards, and their web site
: even includes the current SDMS BIOS along with instructions on how to
: merge it in. Perhaps I should spring for a new motherboard (which
: wouldn't have the silly RZ1000 bogosity). Intel support, do you even
: bother listening to your customers? Or is the retail market something
: you don't care about?
No, they don't care about the retail market. These motherboards are
sold to OEMs. Intel only wants to sell to large companies which will
support the product so they don't have to. Remember the Pentium fiasco?
This is exactly what happened. They quickly had to realize that THEIR
name was on the BigExpensiveChipWithTheBug(tm).
End user support is something Intel wants to avoid. It costs a lot of
money.
: I'd gladly help out to the extent I could with this project. Maybe
: it's time to crank up the ol' disassembler on fmup.
I'd love to tackle this, but I just really don't have the time
unforunately. I started, and have all of the flash BIOS header information
in a file somewhere I can dig up. My next step was to tackle their
weak encryption either through a de{*filter*}/disassembler or just looking at
and analyzing it. There doesn't appear to be a key, it looks like an
identity. Maybe something along the lines of: add 76 to a byte and XOR
it with the value of the byte before it. This is most telling when you
see large areas of 00's in memory and the corresponding area in the BIOS
file is filled with FF's (or some other value I forget).
My solution is going to be buying an ASUS or SuperMicro (depending on
what their support looks like) motheboard. Intel makes a great motherboard
if you're happy with the current features and bugs. Be _very_ careful
when buying an Intel product. Make sure you know everything about it
and can happily live with all of the bugs in the hardware and software.
In fact, I'm interested in the Opti Viper chipset since Intel seems to
be abandoning parity memory. Intel seems to be saying, "Either you have
a server and need ECC memory or your data just isn't important." with
their Triton and Triton II chipsets.
Ah well, just had to vent my disgust with Intel's direction.
Chris
