I have to ask, considering the staggering number of problems posted here just over the last several months that I've been watching this forum, even with current boards.
What's going on with this recurring theme? Is it really that delicate a process when done right? Is there a real, working fail-safe, or is that problematic too (I understand that it's not a "DualBIOS," a fully rendundant hot-spare like some others have, but what is its worth)?
I really don't think it's user error in this many cases.
Many manufacturers don't even provide anywhere near the range of options Intel does to update BIOS versions and until I'd owned an Intel board I'd never found a BIOS upgrade that'd run from within Windows as easy as the Express method.
I do think there seems to have been a large number of people affected recently by buggy BIOS updates though - specifically where users had additional hardware such as discrete graphics cards or customised BIOS settings and in one recent case not the correct BIOS version already installed to update from (not that this was specified as a pre-requisite). People were able to recover using one of the various methods or get there boards replaced under the RMA process though. I do think that more testing should have been undertaken before Intel released some of these BIOS updates to weed out at least the last issue although I don't think the BIOS update system is flawed - far from it.
What is your specific issue (with which board) - or are you making a general, non-specific comment?
Not a specifc issue, but a hope that we might be to consolidate what's been happening and determine how effective recovery methods actually are.
Quite a lot of the messages here are of the sort "Flashed, black screen, recovery jumper didn't work," which is distressing. Intel didn't go the Gigabyte route with a redundant BIOS, but what are the conditions where whatever the jumper is doing doesn't work? If there are going to be buggy BIOS's, or ones with undocumented prereqs, then it would be good to nail this down. I don't consider replacing the board an actual recovery method. That's hours of work and days of downtime.
So just to clarify, your own PC(s) is/are unaffected? In the example you gave there are often other factors at play such as customised BIOS settings/overclocking or additional hardware such as discrete graphics cards installed.
Yes it'd be ideal if Intels BIOS upgrades worked in all situations with all hardware combinations but it'd be an impossible undertaking to try and test for that so the best advise people planning an update could follow would be to simplify their system back to basics as much as possible to remove these 'custom' elements before they start. THEN the chances of anything going wrong will be MUCH reduced. Yes it's potentially quite alot of work to remove graphics cards and restore BIOS defaults etc but then on the other hand people should only be upgrading BIOS versions when the new version addresses specific problems they are experiencing ie if it aint broke, don't try to fix it!
I have a Gigabyte board with their DualBios facility and when I upgraded it after I first got the board (a long time ago), I found the upgrade method was not as easy as my (comparitively newer) DG45ID Intel board has been to upgrade. I believe this technology also adds cost to the board as it means there's additional flash memory required just for the remote possibility of a corrupted flash taking place - to me that's overkill for a home PC and as long as the manufacturer provides good support in the form of an RMA board under such circumstances, I'd rather save money with a cheaper board to start with. Server technology is a different matter however because in that case the ability to quickly get things up and running again must take priority over slightly dearer initial purchase prices.
I think if a little more testing by Intel took place on their BIOS upgrades before release and people reverted to basic systems before updating (and updated only when the BIOS upgrade would fix a known issue they were experiencing, and they followed the instructions precisely), the vast majority of these problems would be gone.
So far, yes, but I'm none too anxious to upgrade ever again even though I'd actually like the "updated VBIOS" or whatever else seems important/relevant, and
these things will eventually come along for a late-2011 board.
You're right, I'm sure DualBios does add cost, and it's certainly most important with server boards, but Intel's high-end enthusiast boards are in several cases as expensive or moreso than the competition ($200+). They're already priced for such a feature and should have it.
If that's the ideal, I think it's still an open question how resilient what we do have is by comparison. I'm a little unclear as to what Intel's recovery method does and doesn't get you out of.