BIOS Software Coming to the End of the Road

I'm sitting here thinking - say we do move to UEFI in the near future...it's a point and click, "user friendly" sort of deal, right? Right...there's still gotta be some sort of underlying, text-based system that the average user will find complicated, that only techs or "computer-smart" people will know how to deal with, right?
 
Ryokurin said:
Some companies such as Phoenix Technologies have been fighting it in favor of their own system, others like Gateway and MSI did it on a few systems but basically only emulated a normal bios but they probably couldn't do much more than that because Windows didn't officially support until Vista Sp1/WS2008, and even then only in 64-bit windows. My guess is now that 46% of new Windows 7 machines are running 64-bit more people are willing to take a chance on it.
Click to expand...

Phoenix Technologies is a has-has-been company.

But even if every single new system and motherboard in 2011 used EFI, what about all the video and other add-in cards that are on the market now, and for the past N-years since PCI-E and PCI were introduced?

Aren't we going to see the convention BIOS as a "subsystem" within EFI for many years?
 
Many (if not all?) of HP's modern laptops -- or at least the business ones -- have UEFI firmware. Unfortunately, on my laptop (EliteBook 8530w), using UEFI boot mode results in many things being broken, both in Windows* and Linux**. I've heard that the new ones are better.

UEFI can be fun to mess with -- I once loaded Apple's Firewire modules, and was able to UEFI-boot Linux from Firewire even though my system doesn't normally offer firewire booting.

I believe HP still defaults to booting with the Compatibility Support Module (CSM), which uses the legacy MBR boot method.

Oddly enough, their desktops (even the business models) do not use UEFI.
* http://social.technet.microsoft.com...g/thread/fefe7c7b-8c9d-472f-9589-327d32f30b06
** https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/612432
 
So, is this new system software based? Where is it stored?
 
What real advantage does UEFI bring to non-servers?

1) Boot time?

Who cares? I can hardly remember the last time I booted any of my computers. S3 powers on instantly. Even if I didn't use it, booting is something that would have to be done once a day at most. Computers are for using, not booting constantly.

2) 2 TB drives

If I understand correctly, the 2 TB limitation is for boot PARTITION only. Who needs a 2 TB boot/system partition anyway?

3) there are potential security risks

4) OC-ing might be more problematic

5) many find it to be yet another overly complex layer


So, is there any real advantage for most users?
 
It's great for accessibility too: Devices like touch screens or whatever hardware exist for people with disabilities can easily be recognized as input devices just like a keyboard.
Internet access will probably also open up a lot of rich features. The wikipedia page about UEFI mentions none of these, so there may be other interesting features too.

But it will take quite a while to provide and debug a full BIOS equivalent, with all the tools like safe flashing of new UEFI revision. The manufacturers who are early in the game will have a real advantage over competition. I've seen some Asus desktop motherboards with UEFI too, for instance. Except they don't advertise them as UEFI, but under the cooler name "Express Gate" and it comes with Skype, MSN Messenger, a web browser, etc. My current BIOS-based Asus motherboard can already act as a skype phone server base even while the PC is shut down, so I guess UEFI will make that even easier. I can imagine new PC cases in which my currently totally useless front LCD display would then say "You have 4 IMs waiting" or something. Microsoft introduced such small second display panels mostly for laptops a few years ago, but it never caught at the time.
 
BTW Asus "Express Gate" is the rebranded Linux distro "Splashtop", not just a UEFI implementation. I have no idea if UEFI is part of Splashtop or a separate thing, as I couldn't find much information on it.
 
Another example of what UEFI lets you do: preboot authentication via fingerprint or smart-card -- with multiple users for the same machine. You can then set the system to automatically lock and unlock (with ATA security) the hard drive, so only the BIOS administrator has access to the real ATA owner password.
 
Meeho said:
What real advantage does UEFI bring to non-servers?

1) Boot time?

Who cares? I can hardly remember the last time I booted any of my computers. S3 powers on instantly. Even if I didn't use it, booting is something that would have to be done once a day at most. Computers are for using, not booting constantly.

2) 2 TB drives

If I understand correctly, the 2 TB limitation is for boot PARTITION only. Who needs a 2 TB boot/system partition anyway?

3) there are potential security risks

4) OC-ing might be more problematic

5) many find it to be yet another overly complex layer


So, is there any real advantage for most users?
Click to expand...

Wow, you sure convinced me. You don't need it, so nobody else does.
 
chinesepiratefood said:
Wow, you sure convinced me. You don't need it, so nobody else does.
Click to expand...

I asked for real benefits and got good replies. Faster boot time and >2 TB boot partition, yes, I dare to say, are not really useful and/or important for majority of people.
 
Meeho said:
What real advantage does UEFI bring to non-servers?

2) 2 TB drives

If I understand correctly, the 2 TB limitation is for boot PARTITION only. Who needs a 2 TB boot/system partition anyway?

So, is there any real advantage for most users?
Click to expand...

Not partition but drive. With a conventional BIOS you can only have a 2TB drive since you need to use MBR which doesn't support anything larger. A conventional BIOS (as far as I'm aware of) cannot boot a GPT drive which has a MUCH higher limit.

Any 64 bit version of Windows supports GPT for any drive EXCEPT the boot drive. (Vista and 7 can boot from GPT with EFI).

Any 32 bit version of Windows since Server 2003 (not XP) supports GPT for any drive EXCEPT the boot drive. No 32 bit version of Windows can boot from a GPT drive EFI or otherwise.
 
Meeho said:
I asked for real benefits and got good replies. Faster boot time and >2 TB boot partition, yes, I dare to say, are not really useful and/or important for majority of people.
Click to expand...

when you put it that way, lets get rid of the core i7 processors, Discrete graphics (as much as we think this is big, we're less than 10??%), sound cards, 3rd party heatsinks and watercooling supplies.
 
I would buy a new motherboard tomorrow if I could get an AMD board with UEFI. AFAIK, there are no AMD motherboards with UEFI.

This is prime time to change, with both companies launching new sockets next year. I am sure Intel, AMD, and the motherboard companies are all being urged by Seagate and WD to roll out UEFI sooner rather than later. The longer they sit on those 3TB+ drives, the faster they lose ground to SSD's. Microsoft is probably in on this as well. As a programmer, I think they are probably itching to drop tons of legacy code and drivers from the next OS or the one after that.

Every new system I build, I still hate to see things like PS/2 ports, floppy ports, IDE ports, serial headers, PCI slots, etc. Getting rid of the PCI bus entirely will be the biggest part of going legacy free. Even then, the motherboard manufacturers will probably keep using add-on controllers to re-enable whatever AMD and Intel decide to drop.
 
Boot times will always be long as software will always have to be loaded into memory. That process will always take time.
 
Ryokurin said:
Some companies such as Phoenix Technologies have been fighting it in favor of their own system, others like Gateway and MSI did it on a few systems but basically only emulated a normal bios but they probably couldn't do much more than that because Windows didn't officially support until Vista Sp1/WS2008, and even then only in 64-bit windows. My guess is now that 46% of new Windows 7 machines are running 64-bit more people are willing to take a chance on it.
Click to expand...

Right there is why we have the reason the BIOS has not already been replaced by UEFI.

Corporations still want to be able to use Windows XP. All major computer manufacturers want to keep their corporate markets.

If the high volume big box manufacturers don't get on board, there is less of a volume to spread the UEFI investment over, so enthusiast level boards are less likely to come with it either.

I see the fate of the BIOS tied to the fate of Windows XP. The sooner we can convince corporate idiots still running obsolete business apps that may or may not work in Windows 7 or Vista to abandon XP, the sooner we will see computers without a BIOS.
 
Conker said:
Boot times will always be long as software will always have to be loaded into memory. That process will always take time.
Click to expand...

Agreed, but if you cut out the BIOS time from the equation, the boot time goes down significantly.

My Machine - currently - spends more time in BIOS & Post mode after hitting the power button than it does actually booting windows / linux.

If I didn't have to wait for the bios, I could be up and running from a cold boot with a functional desktop in less than 15 seconds.

Granted I have a pretty fast SSD, but still...
 
POST time doesn't really seem to be a universal issue. I just built a Media Center PC using an Asus M4A785-M motherboard, and it runs through POST so quick I have trouble hitting the Delete key to get into the BIOS.

Power button, Asus logo flashes for a moment, then I get the Windows 7 bootscren.

Granted, that's a PC with minimal hardware. Even so, I see no reason why we need EFI to gain faster boot times.
That's not to say there aren't advantages to using EFI...i just don't think it's necessary for instant-boot.
 
Meeho said:
Who cares? I can hardly remember the last time I booted any of my computers. S3 powers on instantly. Even if I didn't use it, booting is something that would have to be done once a day at most. Computers are for using, not booting constantly.
Click to expand...

I never use suspend, standby or sleep modes, other than on laptops. My rigs go down for a complete shutdown every time they are turned off, and boot up from a cold every time they go up.

This is probably because I remember how unstable this used to be, and I simply don't trust it. No better way to avoid memory leaks, slowdows and instability than to completely shut the computer down every time you are done using it.
 
Well UEFI does not cut out BIOS, it replaces it with a different software. For that matter, add a second or two because the main objective of UEFI is to make it easy to write drivers for every device and internal component, so it's written in C that's slower to execute rather than optimized Assembly.

But I don't mind, if that means more supported devices right at this UEFI level rather than cumbersome OS-specific drivers defining everything, including the low-level I/O interface.

It will probably take a CPU faster by 100-200MHz to make up for the loss of time induced by UEFI, kind of the same thing that happened when games started using DirectX libraries instead of low level graphic card features. But we all benefited from that, right? Faster development cycles, more and richer features, fun bugs to investigate, and so on.

I know I personally won't use my 3TB drive as a boot drive in a normal usage, but I still want to be able to boot from it in a snap, for instance as a temporary boot drive in case of a main drive failure. But mostly, all these BIOS MBR/EBR obsolete notions of heads and cylinders and prehistoric size limits for partitions must go, the partition scheme must be more dynamic and catter for headless drives such as SSDs.

rclapcar said:
Not [boot] partition but drive.
Click to expand...
The 2.2TB (2TiB) MBR limit is actually for both the boot drive and boot partition, and as far as boot, not disk usage, is concerned, it's really the partition limit that's the blocker.

Still, I can't help wondering if it's a real limit or just a pretext to force us to accept the UEFI GPT new scheme: As far as I know, the MBR or rather EBR limit is 2^32 sectors, but with current 4KiB sector sizes, like Seagate's 3TB drive's, wouldn't the maximum partition size be automatically pushed to 17.6TB (16TiB)?

Also, it means that we could still use the first 512 bytes of the MBR for standard MBR/EBR compatibility, and the next 3.5KiB of the same sector to support zettabyte disks, RAID schemes, non-conventional storage devices (I'm sure a tin of Marmite has far more storage capacity than 3TB), etc.

I really don't see the problem with the existing Windows XP machines, UEFI applies to new UEFI machines only, there will still be BIOS-based machines for some years, or some BIOS/UEFI mixed machines for compatibility, and inventory tools for admins will still support BIOS-based machines.
 
You must log in or register to reply here.
Back
Top