BIOS updating

ubuysa

The BSOD Doctor
This is from a respectable site and yet IMO it's completely and utterly wrong.

https://www.xda-developers.com/please-for-the-love-of-all-things-holy-update-your-bios/

It seems to me to have been written by someone who has just discovered how to flash their BIOS and thinks they need to tell everyone else what fun it was. Many of us on here, and in the other fora to which I contribute, know perfectly well from long (many decades in most cases) experience that only a very tiny percentage of user's problems end up being caused by an out-of-date BIOS. And I do mean a tiny percentage.

It is true that there are times when a BIOS update is important. The recent Intel i9-14900K and i9-13900K voltage regulation issue is a case in point. But these BIOS updates are communicated widely and are applicable to everyone. I honestly cannot remember the last time I saw a problem that was fixed by updating the BIOS. There are a handful I'll grant you, but not many.

Most end user problems are caused by the end user. Either misconfiguring software and/or hardware, installing the wrong drivers, not installing drivers at all, messing with the hardware when they don't know what they are doing, and by blindly following any and all advice they find on the Internet - much of which (like this article) is just plain wrong.

All of us on here, when diagnosing a problem focus first on those things that are the most obvious cause, based on the problem information we can see. We then peel away the layers, like an onion skin, slowly drilling down into less likely causes, again based on addition problem information that we may ask for. An outdated BIOS is at the very (very!) bottom of that list, simply because it's rarely the root cause of problems. Blindly updating the BIOS as the first thing you do, based on zero evidence, is what we call poke-and-hope, and that's no way to go about troubleshooting.

If your car won't start in the morning you don't, for example, immediately update the engine management software. It's far more likely that you have no fuel, a bad battery, a clogged air filter, etc. etc. It's the same with PCs - you always look for the obvious causes first, based on the information you have. And that's almost never the BIOS.

I'm not saying that you should never update the BIOS of course, but suggesting it should be the first step in problem determination, as the article says here...
The moment something is off with your PC, though, a BIOS update should be the first thing you reach for.
...is just plain wrong.

Opinions?
 

SpyderTracks

We love you Ukraine
I’ll preface this by saying personally, I would never flash a BETA BIOS, the timeline between a BETA and public release is usually so small anyway, the times where it’s so essential to do so are almost meaningless. I’d very much stick to this with the AMD and Intel voltage corrections. There are very real risks with a BETA BiOS

While I know this is not going to be favoured, I do understand where that person is coming from.

Times have changed enormously, those in the camp of “if it ain’t broke, don’t fix it” are from a time that updates weren’t applied so much over the internet for software and system firmware. And also from a time before BIOS Flashback capabilities or secondary backup BiOS chips, let alone very stable windows flashing tools. Most BIOS are now automatically served through windows update also. And you can now downgrade a BIOS on most boards which you never could before so the idea of bricking a board from a failed flash is so much more difficult these days, and that was the entire reason for so much caution in the old days

On custom laptops, they’re still quite low level in the flashing capabilities so I treat those entirely separately.

There are so many variables these days that can lead to system instability where a BIOS update is the required

The recent 24H2 update for instance resulted in instability only publicused as affecting Asus boards which was a nightmare to troubleshoot.


Also, Zero Day uncovering is not like it used to be when we unearthed a serious one perhaps every couple of years, nowadays you get multiple regularly.

Just look at this dell device as an example of how many critical BIOS updates one after the other they’ve had to release, many of these I know full well are related to 24H2 (having dealt with this specific chassis for clients), but you’d never know that as Dell haven’t reached out in anyway to notify users and Microsoft haven’t publicly referenced it like with Asus, and as with all BIOS updates, the notes on what it addresses is just shocking, and these critical updates and poor notes are pretty much across the board for all manufacturers


With the current reworking of the windows kernel that Microsoft are doing after the CloudStrike outage last year, you’ll find BIOS updates will be far more routine for a while until they get it stable, and since the last 3 month windows cumulative update packages have resulted in BSODs related to Kernel Panic, they’ve got quite a way to go.
 
Last edited:

SpyderTracks

We love you Ukraine
I’ll preface this by saying personally, I would never flash a BETA BIOS, the timeline between a BETA and public release is usually so small anyway, the times where it’s so essential to do so are almost meaningless. I’d very much stick to this with the AMD and Intel voltage corrections. There are very real risks with a BETA BiOS

While I know this is not going to be favoured, I do understand where that person is coming from.

Times have changed enormously, those in the camp of “if it ain’t broke, don’t fix it” are from a time that updates weren’t applied so much over the internet for software and system firmware. And also from a time before BIOS Flashback capabilities or secondary backup BiOS chips, let alone very stable windows flashing tools. Most BIOS are now automatically served through windows update also. And you can now downgrade a BIOS on most boards which you never could before so the idea of bricking a board from a failed flash is so much more difficult these days

On custom laptops, they’re still quite low level in the flashing capabilities so I treat those separately.

There are so many variables these days that can lead to system instability where a BIOS update is the required

The recent 24H2 update for instance resulted in instability only publicused as affecting Asus boards which was a nightmare to troubleshoot.


Also, Zero Day uncovering is not like it used to be when we unearthed a serious one perhaps every couple of years, nowadays you get multiple regularly.

Just look at this dell device as an example of how many critical BIOS updates one after the other they’ve had to release, many of these I know full well are related to 24H2 (having dealt with this specific chassis for clients), but you’d never know that as Dell haven’t reached out in anyway to notify users and Microsoft haven’t publicly referenced it like with Asus


With the current reworking of the windows kernel that Microsoft are doing after the CloudStrike outage last year, you’ll find BIOS updates will be far more routine for a while until they get it stable, and since the last 3 month windows cumulative update packages have resulted in BSODs, they’ve got quite a way to go.
Just to follow on from this, the reference to CrowdStrike and Microsoft reworking the Kernel, this is why all these BSODs have been happening since 24H2, it’s down to them being mandated by pretty much everyone on the planet to lock out unchecked Kernel access like on security products


This is also why they’ve chosen now after many years of allowing it to not renew the signing of the WinRingZero driver, it’s because worldwide regulators are forcing them to to avoid these major system outages


Its also why 24H2 still isnt available on systems with certain games that use kernel level anti cheats, although those things should be illegal in the first place and aren’t attached to any decent games so I have no sympathy there
 
Last edited:

ubuysa

The BSOD Doctor
I have no argument with any of that. My objection is the assertion that a BIOS update should be the first thing you do when you have problems. That's just plain wrong.
 

SpyderTracks

We love you Ukraine
I have no argument with any of that. My objection is the assertion that a BIOS update should be the first thing you do when you have problems. That's just plain wrong.
But that's where I differ nowadays.

From the days of troubleshooting 24H2 issues and eventually finding the cause was an outdated BIOS, not highlighted at all on the vendor update site. It was needless time wasted.

IF there were still the old problems and risks with updating BIOS, IF there were clearer mandated bullet point update notes for all manufacturers so you could unquestionably pin-point what they addressed, then I would agree.

But with the advances in both BIOS update frequency, windows Kernel rewrites AND flashing stability and redundancy, it's just a routine update you can include with driver updates to further be sure there are no firmware / software variables impacting troubleshooting.

The question is no longer why you should but why wouldn’t you?

For me it’s about prevention rather than cure. I find if I keep things updated, I rarely have to troubleshoot.
 
Last edited:

ubuysa

The BSOD Doctor
Perhaps I'm old school, but I much prefer a logical and step-by-step approach to troubleshooting based on real data. This "just update the BIOS and hope for the best" doesn't wash with me and never will. If the data suggests a BIOS flash is wise then that's what I'll suggest, but in my experience (even recent experience) it's almost never the BIOS that's the root cause.

I do agree with your "why wouldn't you" statement, and my BIOS is up to date, and will stay up to date - although I still research the latest version as much as I can first. But that's a world away from "The moment something is off with your PC, though, a BIOS update should be the first thing you reach for"...
 

SpyderTracks

We love you Ukraine
Perhaps I'm old school, but I much prefer a logical and step-by-step approach to troubleshooting based on real data. This "just update the BIOS and hope for the best" doesn't wash with me and never will.
But that's not what you're doing. You're prepping the firmware / software layer to reduce possible variables that could confuse results when troubleshooting, if you're really lucky it may well give a fix as a result. The point is by not doing it, you increase the variables that could be a factor to the issue, and needlessly spend extra time either trying to troubleshoot an issue the BIOS addresses, or potentially adding symptoms to the issue leading to confused avenues of investigation.

It's not to give any kind of fix, simply to prepare the OS level as much as possible so that hopefully any unknown future instabilities may be side stepped, as you're already up to date.

The reasons for not doing it that have been drummed into most of us who are a bit older aren't a factor anymore. I totally get it, believe me, my initial instinct is still "if it's not broke don't fix it", but (moreso on laptops), I've found quite a number of instances where BIOS has been the fix.

I'm like you though, I've never had a failed board, never had a corrupt BIOS, not had a failed BIOS flash and on my own personal devices, I've never had a BIOS update correct a current issue (other than AGESA/Microcode updates, but those are different as you say). I don't know if I'm simply lucky, or a little of it is down to significant research into quite high quality components that may reduce issues, or what that particularly means.
 
Last edited:

ubuysa

The BSOD Doctor
But you're making my point in that latest post. In my experience, and not just from my helping but from posts on many fora solved by others, the percentage of problems we see that turn out to be fixed by a BIOS update are tiny - far less than 1% even. Given that, suggesting that the first thing you should do when you have a problem is update the BIOS is contrary to the evidence!

Again, I'm not saying don't update the BIOS - I do, though only after the update has been available for a month or two and I've researched it as much as I'm able. What I am saying is don't update the BIOS as the first thing you do when you have a problem!!
 

SpyderTracks

We love you Ukraine
But you're making my point in that latest post. In my experience, and not just from my helping but from posts on many fora solved by others, the percentage of problems we see that turn out to be fixed by a BIOS update are tiny - far less than 1% even. Given that, suggesting that the first thing you should do when you have a problem is update the BIOS is contrary to the evidence!

Again, I'm not saying don't update the BIOS - I do, though only after the update has been available for a month or two and I've researched it as much as I'm able. What I am saying is don't update the BIOS as the first thing you do when you have a problem!!
Kind of missing the point of what’s being said though, you're still coming at it from a perspective of a reason to be hesitant in upgrading, but haven't said what that threat is?

Again, why wouldn’t you update it before troubleshooting in the same way you would a driver? The reasons for that hesitance is based on outdated risks that are no longer present, if it doesn’t change anything, then what does it matter if it’s updated or not?

The amount of live issues it resolves is irrelevant in the same way as updating a driver.

The relevance is the possibility of limiting variables that you’re unaware are related.

You’re looking at the 99% being needless rather than the 1% where it took x number of hours to identify a cause simply because they were still adhering to old risks that are no longer relevant.

I would also say, certainly on laptops, since the release of 24H2, the number is far higher than 1%
 
Last edited:
Top