November 10, 2011, 3:42 PM —
flickr/Enrique Vicente
When you turn on your computer, a primitive system that dates back more than 30 years, the basic input/output system (BIOS), turns your cold hardware into a functioning system that your operating system can then boot from. Alas, it's sadly out of date. PC makers have slowly been replacing BIOS with the Unified Extensible Firmware Interface (UEFI). That's all well and good, but one UEFI feature, Secure Boot, could be used to lock PCs into being only able to boot one operating system: Windows 8.
[ Linux Foundation recommends fixes for UEFI roadblock ]
So, what's really going on here? Is UEFI just a way for Microsoft and its most loyal original equipment manufacturers (OEMs) to keep Linux and other alternative operating systems out or is it more than that? To answer that, let's take a look at what's what with UEFI.
What is UEFI
If you're in the computer hardware business, you know that BIOS has been terribly outdated for decades. For example, a BIOS only has 1,024KB (kilobytes) of executable space. That, in turn means, a BIOS has trouble starting up the multiple peripheral interfaces (USB, eSATA, ThunderBolt, etc.) devices, ports, and controllers on a modern PC. Just as annoying, the BIOS was never meant to initialize more than a handful of devices so even if you can get all devices ready to go it will take up to 30 seconds after you turn the switch on before your PC is ready to start booting.
The computer companies aren't dumb. They knew that BIOS was obsolete even before the 21st century dawned. But, until recently they couldn't agree on how to replace it.
In 1998, Intel started work on the “Intel Boot Initiative” (IBI), later known as Extensible Firmware Interface (EFI). While Apple, in its Intel-based Macs, and HP, with its Itanium 2 servers, used it, the other OEMs and, needless to say, Intel's rival chip vendors, weren't initially keen on adopting EFI. In 2007, Intel, along with AMD, AMI, Apple, Dell, HP, IBM, Lenovo, Microsoft, and Phoenix Technologies, finally agreed to use UEFI (the re-branded EFI) as the universal replacement for BIOS.
Don't mistake UEFI as being purely a BIOS replacement. It's not.
UEFI is a mini-operating system that sits on top of a computer's hardware and firmware. Instead of being stored in firmware, as is the BIOS, the UEFI code is stored in the /EFI/ directory in non-volatile memory. Thus, UEFI can be in NAND flash memory on the motherboard or it can reside on a hard drive, or even on a network share.
Even in a UEFI system there will still be a little bit of the BIOS in the firmware to enable UEFI itself to "boot" up.
The UEFI advantage
The first thing you'll notice about UEFI systems is that they boot faster and you can have even larger primary drives. The BIOS is unable to boot from hard disks with more than 2.2TB (terabytes). That's a hard limit set in the Master Boot Record (MBR) that you can't fix. In the BIOS MBR, the maximum space for a drive is determined by the formula: 2 to the 32nd times 512 bits. This is an old hard drive addressing scheme. What it means in practice is that all but the most up-to-date computers can't boot with hard drives that are larger than 2.2TB. With 3TB drives now becoming common, OEMs have no choice but to move to UEFI on high-end PCs.
UEFI uses the GUID (Globally Unique ID) Partition Table, both to replace the MBR and address partitions. With GUID, you'll be able to boot from hard disks as large as 9.4ZB (zetabytes). How big is that? Well, everything -- and I mean everything -- on the Internet is believed to be just over 3ZBs. I don't think we have to worry about UEFI not being able to manage any drive it's likely to run into anytime soon.
