The end of the old PC as we know it?

One of the last and least-loved remnants of the original IBM PC, the BIOS, is about to get its marching orders, Intel predicts.
Written by Rupert Goodwins, Contributor
SAN JOSE--One of the last and least-loved remnants of the original IBM PC is about to get its marching orders, Intel predicted at its Developer Forum here this week.

For more than 20 years, the PC has relied on the BIOS (Basic Input/Output System), a small set of fixed software routines normally built into a chip on the motherboard. This hangover from a distant past is causing more and more problems, said Mark Doran, Intel's principal engineer behind the Extensible Firmware Interface (EFI) that aims to humanely kill the antique technology.

"When I started, I got senior managers together from across Intel and asked them what would happen if we had a blank sheet of paper to replace the BIOS," Doran said. "It turned into a three-day bitch session." He said that the original designers of the IBM PC BIOS had no idea that it would survive this long. "They thought that 250,000 machines would see it through to the end of its life," he said.

Most people only know the BIOS as the mysterious source of arcane configuration and test messages that appear when a PC is first switched on. Although its job is to connect the various bits of hardware in a PC to the operating system in a standard fashion, there are no standards for how it is created or configured--and it's not uncommon for extension boards from different manufacturers to clash in their use of the BIOS.

And if people do use the BIOS to change configurations or diagnose a problem, there's often not much help and few similarities between different PCs. "Even experienced support staff often end up in 'Now what?' mode with the BIOS," Doran said.

EFI promises to change all that. EFI is a tiny operating system in its own right, freed from the constraints of the BIOS. The first difference people will see is in the splash screen, the display the PC shows when first turned on. Unlike the BIOS, EFI supports high-resolution displays. Likewise, it can run with a proper graphical user interface, rather than the blocky text-only interface. EFI also has its own networking, so it can be used for remote diagnostics.

The differences are much more than just the interface, Doran said. "The BIOS is the last place on the PC where people have to write in low-level assembler code, and we want to end that," he said. Instead, EFI is almost entirely written in the C programming language and allows additions to be created using standard programming tools. Such additions can include much more detailed and useful diagnostics, self-configuration programs, and ways to sort out problems even if the operating system has died.

"We even have a concept of the afterlife, so if your OS freezes you can go in and look at the state of the machine, change the configuration, load a different driver, and do a sensible restart," Doran said.

As part of the demonstration, he showed a network driver being replaced on a live machine, as well as multiple reconfigurations of various USB (universal serial bus) devices. Because EFI has its own filing system that lives on a reserved part of the hard disk, it can become the standard home for a whole set of utilities that have always had an awkward fit with the BIOS. Digital rights management and security designers also have an interest in EFI because it gives them a new level of control over the hardware.

Finally, EFI can pretend to be a BIOS. "We're not expecting people to throw out the BIOS overnight, so EFI can support legacy systems by running on top of an existing BIOS and handing over control when appropriate," Doran said.

ZDNet UK's Rupert Goodwins reported from San Jose.

Editorial standards