Will BIOS Be Dead in 3 Years?

[xxsk8er101xx]

hopefully BIOS is friggen old lose it already.

 

[Tamz_msc]

Sigh!Even though technology must move forward, I'm gonna feel sad for BIOS because fiddling with it makes me feel like a geek.But anyway,its my opinion and I definitely want UEFI to move forward and replace BIOS but it won't be in 3 years.

 

[liquidsnake718]

I think the bios should still be an option, like DOS

 

[kingssman]

http://docs.hp.com/en/5991-1247B_ed2/ch04s13.html#aes-npar-287b

A link to potential EFI control commands to set PCI ports, memory, load/unload drivers ect.

 

[kingssman]

[citation][nom]7amood[/nom]as long as I can control my CPU multiplier, clock and voltage... bring it on...[/citation]
Well imagine this, tweaking your CPU multiplier via terminal command inside windows or linux OS and be able to overclock w/o reboot. That's the potential of EFI

 

[oxxfatelostxxo]

correct me if im wrong, but i thought it limited the ability to read any hdd over 2tb, whether storage or boot drive?. (if so, then that would be a issue in time, though not many of us use over 12tb+ of total hdd (2tbx6 sata ports or more) for those that just got confused

If its just the boot drive then it really doesnt make much diff to me. since i for the most part only run the OS on the main drive.

Not to mention we are taking a step backwards in size with ssd's being the hdd replacement for main drives.

 

[ta152h]

This technology, or something very similar, was introduced in 1986 on the IBM RT PC.

They should get rid of Real mode, Virtual 86, and Protected 286 mode as well. Does anyone still run DOS apps on a modern processor? I can't imagine it's many, and everyone else is paying for the extra silicon, and extra power use. With BIOS gone, the best reason for keeping Real Mode is gone.

 

[kronos_cornelius]

I hope this does not increase boot time. It feels like we are loading more and more intermediary layers to load an OS. UEFI, boot-loader (grub), and the the OS. I am at the point where the main OS (Kubuntu) is the one that takes the least time of all the steps.

 

[ta152h]

[citation][nom]Razor512[/nom]doesn't bigger generally = more data which generally = longer load times and more resource usage? So wouldn't a bios load faster?[/citation]

Not really. If they do it like the IBM RT PC of 1986, which is far more advanced in most ways than a 2010 PC, you essentially free the OS from having to worry about drivers and such. There's a layer that the OS runs on that handles this, which would make the Linux dorks pretty happy.

We're so used to device drivers being OS specific, that we sometimes forget that it makes no sense at all. It should be hardware specific, and the OS should request it from a lower level that handles all that kind of stuff.

Again, this was done in 1986 on the IBM RT PC. It's not new, and it's not very difficult. It makes the most sense too.

 

[techguy378]

[citation][nom]kronos_cornelius[/nom]I hope this does not increase boot time. It feels like we are loading more and more intermediary layers to load an OS. UEFI, boot-loader (grub), and the the OS. I am at the point where the main OS (Kubuntu) is the one that takes the least time of all the steps.[/citation]
[citation][nom]nukem950[/nom]Well, it is about time. I so want an UEFI board.I wonder when an AMD board will come out...[/citation]
UEFI drastically reduces boot times. Also, UEFI motherboards have existed for quite some time. I've successfully done an EFI install of Windows 7 on both an Intel DP35DP and DP43TF motherboard. In the case of the Intel DP43TF motherboard boot time with Windows 7 was cut in half compared with a traditional BIOS install.

 

[ceteras]

[citation][nom]TA152H[/nom]Not really. If they do it like the IBM RT PC of 1986, which is far more advanced in most ways than a 2010 PC, you essentially free the OS from having to worry about drivers and such. There's a layer that the OS runs on that handles this, which would make the Linux dorks pretty happy. We're so used to device drivers being OS specific, that we sometimes forget that it makes no sense at all. It should be hardware specific, and the OS should request it from a lower level that handles all that kind of stuff. Again, this was done in 1986 on the IBM RT PC. It's not new, and it's not very difficult. It makes the most sense too.[/citation]

Interesting, but how does EFI handle drivers for add-on cards? I guess drivers would still be required for them.

 

[welshmousepk]

Couldn't they have given it a name thats easier to say?

 

[Prescott_666]

"Does anyone still run DOS apps on a modern processor?"

Yes. A lot of the bootable Diagnoses and Repair programs run on FreeDos. A lot of them do run on Linux now, and if Dos mode went away, some of the DOS based repair tools would be rebuilt to run on Linux, but a lot of them would not.

As long as the BIOS/EFI supports boot from USB floppy drives, I don't want to give up DOS.

 

[gamerk316]

[citation][nom]TA152H[/nom]This technology, or something very similar, was introduced in 1986 on the IBM RT PC.They should get rid of Real mode, Virtual 86, and Protected 286 mode as well. Does anyone still run DOS apps on a modern processor? I can't imagine it's many, and everyone else is paying for the extra silicon, and extra power use. With BIOS gone, the best reason for keeping Real Mode is gone.[/citation]

Intel prides itself on perfect X86 backward compatability, and I doubt they will drop real mode anytime soon. Besides, I still have a DOS bootdisk ready for when Windows REALLY screws up.

Just last month, due to a Windows update plus a borked ATI driver update, I went into safe mode before doing a re-boot to finish updateing a windows update, which caused Windows to bork and ALWAYS boot into safe mode, regardless of the boot option [which was a problem, as due to the aforementioned Windows Update not being fully installed, Safe mode would try and fail to undo the changes the Update made, causing Safe Mode to toally lock up.

To fix the problem, I had to boot into DOS and modify the registry value that handles which mode Windows boots in, forcing a restart in normal mode [where Windows thankfully was able to fully abort the update].

As an aside, if ATI made an install package that didn't CRASH, I wouldn't have had this issue in the first place...

 

[huron]

[citation][nom]thebob654321[/nom]apple's had efi for how long? yeah it ain't that hard. mobo makers can sell a usb dongle efi to circumvent the bios.[/citation]

Wow...that's quite interesting. I thought that Apple was using Intel boards and chips, which should have all the same hardware specs.

[citation][nom]kelemvor4[/nom]Hogwash.BIOS uses Master Boot Records (MBR) to boot from a drive. The MBR is 512 bytes of which 64 bytes is the partition table, 16 bytes to fully describe each partition's attributes, so you only have 8 bytes to point at each partition's location and size. Only 4 bytes of that can be used to point at each partition's beginning sector. Four bytes is 32 bits, so you can address 2**32 sectors and (2**32)*512 bytes (since there are 512 bytes/sector) = 2.2 E12 or about 2 TB (maximum addressable disk size and also partition size since 4 bytes is also allowed to describe each partition's size with BIOS and MBR).It's a very cut and dry 2TB limitation baked into BIOS. I discovered this rather painfully with my last build when I tried to use a single 6Tb array. Quite disappointing.Now once you have booted from a 2TB or smaller partition using windows/macos/linux/name your bootloader you can use software to mount a GPT partition that exceeds 2Tb. You still cannot boot from a partition greater than 2Tb. There's nothing to argue about here, try it and find out for yourself.[/citation]

This is true, but don't the drives with 2k sectors help to alleviate this? This is not specifically BIOS related, but HDD partition/MBR related - the BIOS just checks to see where to boot the drive, which is pointed out the MBR - the BIOS doesn't care about the size of the drive, etc, does it?

 

[nukemaster]

As long as i can overclock/tweak, I am happy with it. As for the comment about EFI allowing this to be done in windows, So do many boards with software as well(Asus,Gigabyte and more).

 

[tommysch]

Its about time... We should also ditch every 32bit OS.

 

[Regulas]

I would like to see flash drives come down and use a locked down flash drive for the OS only. This if done right could greatly increase security and help boot speed too. Then use a standard HD for data and programs.

 

[Snipergod87]

I still

 

[kelemvor4]

[citation][nom]oxxfatelostxxo[/nom]correct me if im wrong, but i thought it limited the ability to read any hdd over 2tb, whether storage or boot drive?. (if so, then that would be a issue in time, though not many of us use over 12tb+ of total hdd (2tbx6 sata ports or more) [/citation]
You can use partitions greater than 2Tb with bios for sure as long as you're not still on XP. In fact I think I read they added support in sp3 or something, but I upgraded from xp before sp3 came out. MacOS shouldn't have a problem since mac uses EFI and I'm fairly certain you can have data drives >2TB in Linux as well.

 

[annymmo]

What about Coreboot?
http://en.wikipedia.org/wiki/Coreboot

 

[kelemvor4]

Wow...that's quite interesting. I thought that Apple was using Intel boards and chips, which should have all the same hardware specs.

Intel created EFI as I understand it. I had an Itanium workstation from HP at one point with XP IA64.

This is true, but don't the drives with 2k sectors help to alleviate this? This is not specifically BIOS related, but HDD partition/MBR related - the BIOS just checks to see where to boot the drive, which is pointed out the MBR - the BIOS doesn't care about the size of the drive, etc, does it?

I don't believe the sector size is stored on the disk, rather it's programmed into the BIOS. I know the MBR must be 512 bytes but I suppose if the MBR code section had instructions to use a larger sector size for the rest of the disk it might be possible. MBR: http://en.wikipedia.org/wiki/Master_boot_record


As for the driver concerns. Isn't it only low level drivers that would need to be written to firmware such as the ones that display your bios boot screen today? My assumption was that after performing basic boot operations you could load additional high level drivers as you already do in operating systems today.

 

[ta152h]

[citation][nom]Prescott_666[/nom]"Does anyone still run DOS apps on a modern processor?"Yes. A lot of the bootable Diagnoses and Repair programs run on FreeDos. A lot of them do run on Linux now, and if Dos mode went away, some of the DOS based repair tools would be rebuilt to run on Linux, but a lot of them would not.As long as the BIOS/EFI supports boot from USB floppy drives, I don't want to give up DOS.[/citation]

All of those boot disks could be easily re-written in 386 protected mode. If you did that, every core of every processor would be slightly smaller, and thus cost less, and use less power. The impact of removing it is very minor in a negative sense, and very minor in the savings per processor, but huge in the number of processors that would save because of this. Even if you saved 5 cents per processor, and 5 cents a year on electricity, multiply that by the hundreds of millions of processors that get sold.[citation][nom]gamerk316[/nom]Intel prides itself on perfect X86 backward compatability, and I doubt they will drop real mode anytime soon. Besides, I still have a DOS bootdisk ready for when Windows REALLY screws up. Just last month, due to a Windows update plus a borked ATI driver update, I went into safe mode before doing a re-boot to finish updateing a windows update, which caused Windows to bork and ALWAYS boot into safe mode, regardless of the boot option [which was a problem, as due to the aforementioned Windows Update not being fully installed, Safe mode would try and fail to undo the changes the Update made, causing Safe Mode to toally lock up. To fix the problem, I had to boot into DOS and modify the registry value that handles which mode Windows boots in, forcing a restart in normal mode [where Windows thankfully was able to fully abort the update]. As an aside, if ATI made an install package that didn't CRASH, I wouldn't have had this issue in the first place...[/citation]

I don't know why you'd boot into DOS to do what you say. It seems much harder, unless you're not making the distinction between DOS and a command line. I don't even know how DOS would read NTFS, but I guess they could make an add-on. Either way, there are ways to boot into Windows without using DOS, and being able to do the same exact thing. Actually, I don't know anyone that uses DOS for that, and I don't know why anyone would. I'm guessing you're confusing command line with DOS.

 

[ta152h]

[citation][nom]ceteras[/nom]Interesting, but how does EFI handle drivers for add-on cards? I guess drivers would still be required for them.[/citation]

You always need drivers, per se, it's just not part of the OS. You have a lower level, below the OS, which you would update with a new device. The OS sends requests to this layer. So, if you got a new device, you update this. Especially with virtualization become more common, I don't see how the computer can't evolve this way. You would just need instance of this running, even if you had several operating systems open at once. There would also be better efficiency since this one hardware layer would be able to manage several requests and know about them all, instead of each OS thinking it's the only one requesting hardware and have a virtualization layer without as much intelligence dealing with it the best it can.

 

[killerb255]

If you all want an example of an EFI, look no further than a recent MacBook.