Friday 11 May 2012

UEFI: The Super BIOS

The end of the old BIOS is imminent; UEFI is faster, safer and more efficient. We explain why.



The bluish, confusing menu of the BIOS (Basic Input / Output System) has been a feature of every PC for the past 30 years. It’s function is to detect installed hardware and provide the necessary interface for the boot process of Windows and Linux. The BIOS was developed for PCs in the early eighties, and has remained virtually unchanged in recent years. Since 2000, Intel is working on a new firmware interface, the EFI (Extensible Firmware Interface). 
                             Since 2005, the United EFI Forum is responsible for developing, managing and promoting UEFI specifi cations. Big companies such as Intel, AMD, Apple, Dell and Microsoft develop product in accordance to UEFI standards. UEFI has a faster, more stable, easier-to-use and secure interface. Nevertheless, there are hardly any systems that use UEFI as of now. We show you how your computer will boot up in the future using UEFI. 

---------------------------------------------------------------------------------

" By 2015, more than
90percent of IA*-Systems
will boot with UEFI " -  Vincent Zimmer
                                                                            Principal Engineer
                                                                        Software and Services group, Intel



---------------------------------------------------------------------------------                                                                             


Faster: UEFI integrates the drives
 You can observe the boot process carried out by the BIOS once the PC is switched on. For a few seconds, until the Windows boot screen appears, the BIOS executes the so-called POST (Power-On Self-Test). It’s a diagnostic testing sequence which checks the functionality of the CPU, memory and various input/output devices. The BIOS works at an abstraction level so that the operating system doesn't have direct access to the hardware. However, the correct drivers are loaded as the operating system is loading. Finally, the BIOS scans the master boot record or partition table, which resides in the fi rst sector of a partitioned drive. If the boot loader is present on the active partition, the operating system takes control of the system from the BIOS and starts up. The BIOS has two fundamental weaknesses. Firstly, it is based on 16-bit assembly code and cannot directly address the latest 64-bit hardware, and secondly, there are no set standards for specifications, so manufacturers come up with their own versions. The participants of the UEFI Forum wanted to set this straight. From the outset, each process has been precisely defined. Thus, the boot process or platform initialization (PI) is clearly described in phases. Immediately after powering up the PC, the Pre-EFI Initialization (PEI) is executed, which initializes the CPU, memory and chipset. This is then followed by the Driver Execution Environment (DXE). At this point, the rest of the hardware is initialized. This process saves the time required for booting because UEFI can integrate various drivers that need not be reloaded during booting. Thanks to these drivers, the user already has access to network card, including features such as network booting and remote assistance at the early stage of the boot process. With the graphics processor enabled, a fancy user interface is also presented. 
                            However, biggest time-saving feature of UEFI is the fact that not all the installed hard drives will be scanned for the boot loader, since the boot drive is set during the installation of the operating system in the UEFI. The default boot loader is run without consuming much time searching the drives. 
                            The faster boot time is not the only advantage of UEFI; applications can be stored on virtually any non-volatile storage device installed on the PC. For example, programs and diagnostic tools such as antivirus or system management tools can be run from an EFI partition on the hard drive. This feature will be very useful to original equipment manufacturers (OEM), who can distribute systems with extra functions in addition to the standard EFI firmware stored on the motherboard’s ROM. 

UEFI fully supports 3 TB hard drives
 Due to delayed transition to UEFI over and over again, new hardware could be addressed by the BIOS by means of extensions. The classic BIOS can access only up to 232 sectors of 512 bytes in size, which translates to a total of 2 TB. So the upcoming 3 TB variants of Western Digital Caviar Green and Seagate Barracuda XT won’t be fully compatible with the current BIOS. Seagate uses larger sectors to make the full capacity usable on Windows, butthe BIOS cannot boot from  this drive. UEFI, on the other hand, works with GUID partition table (GPT) with 64-bit long addresses and can handle up to 264 sectors that address up to 9 Zettabyte (1 zettabyte equals 1 billion terabytes). Since Windows Vista can handle GPT disks, booting works with only a few UEFI motherboards in the market, such as the Intel DQ57TM. However, in 2011, Intel will sell more motherboards with UEFI than standard BIOS, followed by all the other manufacturers