The BIOS has been around for more than three decades. What made sense then is a weak link in your computing infrastructure now. It may be time to move on.
As you deal with legacy technology and all its limitations, your IT staff will encounter more and more issues going forward: hardware compatibility problems, cost-intensive troubleshooting, slow boot and resume times, and limited pre-boot tools – just to name a few. Compared to newer UEFI systems, computers running the classic BIOS require a higher maintenance effort and are limited in their capabilities. We’ll get to the problems and benefits later.
There is no doubt: UEFI will replace BIOS eventually. However, many managers, IT personnel, and even enterprise architects are uncertain of UEFI’s origins or what it actually brings to the table. By the end of this article, you’ll have a clear understanding of this emerging standard and if it is worth your company’s investment.
What is UEFI?
From a bird’s eye view, UEFI has the same purpose as your typical BIOS. It’s responsible for initializing the hardware of your PCs and laptops, and then handing full hardware control over to the operating system. However, in terms of pure pre-boot functionality and extensibility UEFI is light-years ahead (see “IT Benefits” below), and it also addresses some basic problems with our aging architecture.
BIOS has been tweaked over the course of 30 years to fit modern hardware, such as multi-threaded and multicore processors, but make no mistake: Its core is still designed for PCs with one simple CPU, the long-dead ISA bus, and memory without a memory controller. BIOS was designed for ancient technology and we are starting to feel the implications of building on top of it.
The industry has tried again and again to standardize a successor to the almost decaying BIOS. All of those standards failed. Then, back in 1998, Intel initiated the “Intel Boot Initiative” (IBI) for the enterprise-class Intel IA-32 and -64 processors and later rebranded the initiative to EFI, Extensible Firmware Interface. There was a need to address the growing pains of BIOS, such as its limited storage and pre-boot manageability.
Intel pushed its efforts by eventually handing over the entire EFI initiative and all its specifications to the newly founded UEFI forum, which to this date consists of some of the leaders of our industry: AMD, American Megatrends Inc, Apple, Dell, HP, IBM, Insyde, Intel, Lenovo, Microsoft, and Phoenix Technologies.
The Unified Extensible Firmware Interface was born. UEFI forum members are responsible for keeping the UEFI specifications up to date to and adapt them to the ever-changing hardware and software requirements. “170 members and growing! […] UEFI platforms crossed 50% of IA worldwide units,” Intel announced at the 2011 meeting.
What’s Wrong with Now?
To understand UEFI’s benefits, you need to be aware of some of BIOS’ major flaws.
First of all, servicing a fleet of company desktop PCs or laptops that are in a non-bootable state (i.e. the OS is “broken”) is tough. Managing a PC over the network without Windows or Linux to boot from is time- and cost-consuming. The IT support staff needs a pre-boot environment that permits basic remote troubleshooting and maintenance tools. If anything goes wrong, you want to push a button and restore the PC to a functional state – from the comfort of your desk.
Second, the BIOS was developed in 1979 and is far beyond out-of-date. Its limited execution space (1024 Kbytes!) and its limited number of addressable devices causes more and more problems in today’s architecture. Your IT systems need to handle an increasing amount of devices. BIOS will fail you at some point – it will simply deny handling the device or may even cause problems in an OS environment.
Speaking of devices: Thanks to today’s vast number of USB ports, PCI devices, or built-in controllers (none of which existed 20-30 years ago), BIOS struggles with initializing them one-by-one. This struggle results in delays of up to 30 seconds before the actual operating system start its first boot sequence!
Furthermore, the BIOS is physically unable to boot from hard disks with more than 2.1 TB. Back in the 80s, that kind of capacity seemed like a fairy tale, but we are being ushered into an era of hard disks with multiple terabytes of disk space. (See Solid State Drives for Desktop PCs and Laptops: Ready for Prime Time?)
Want to invest in 3 TB hard disks? Then there’s no (feasible) way around UEFI; only the new GUID table supports booting from anything bigger than 2.1 TB.
The problem lies within the old MBR (Master Boot Record) system used by most BIOSes. The partition table is limited to 2.1 TB, so your clients can’t boot from these devices – at least not without time-consuming workarounds or specialized tools.
“As we move beyond BIOS, we will need a 64-bit pre-OS environment that takes advantage of the UEFI environment while offering shell scripting ability, hardware access, simplicity and familiarity,” wrote Dong Wei, vice president of the Unified Forum and HP Distinguished Technologist in the official UEFI book, Harnessing the UEFI Shell.
UEFI Benefits and Innovation
UEFI addresses all of the abovementioned issues. However, keep in mind that UEFI is not a single “product” that looks exactly the same on every type of PC or laptop. UEFI is a basic firmware with a huge framework. To make full use of UEFI, the hardware needs to specifically support and come with UEFI applications. It’s a flexible design.
However, many UEFI implementations are highly developed and share common benefits:
- Enterprise management: Thanks to support for third party drivers and applications, your IT staff can remotely manage systems equipped with UEFI without booting into Windows or any other OS! This enables you to have a common infrastructure for managing computers across your entire network. Obviously, this is a money saver: Imagine fixing computers by remotely turning them on and running UEFI troubleshooting tools or restoring the image from a backup server. It essentially eliminates the need for the classic “system DVD.”
- Pre-OS and network security: UEFI’s enhanced networking API allows for a rich network authentication (log-on) in a pre-boot environment. Also, it offers support for TPM and and authenticode signatures. It’s simply an additional layer of security.
- UEFI is “OS-like:” By default, you have full access to the entire hardware of your computer – Ethernet adapter, Wi-Fi, Bluetooth, graphics card, USB, even the audio chip, as well as full blown x86 and x64 support. This enables not only high-resolution UIs, but highly functional pre-boot environments. HP put its System Diagnostics tool (written specifically for UEFI) on laptops and desktops starting in mid-2008.
MSI puts ClickBios on its board, a UEFI environment with some basic maintenance, diagnostics and Instant-On environments (for gaming, multimedia etc.). Unfortunately, that MSI mainboard – based on the P45 chipset – has been discontinued.
Note, these interfaces aren’t mandatory. Some manufacturers may build beautiful UIs in their UEFI environment; some may just skip the UI aspect entirely.
- Faster boot and resume times: UEFI handles devices initialization within seconds. This increases IT staff productivity (and user impatience), especially if your business requires rebooting or going in and out of hibernation a lot.
- Support for HDDs with more than 2.1 TBytes: UEFI solves this problem by introducing the newer partition table called GUID (global unique identifier), which is finally capable to address more than 2 terabytes of storage. Bottom line: If your business revolves around having huge amounts of data on your PC client, investing in UEFI-capable hardware is a no-brainer.
- Specialized UEFI applications: Without the need to boot into an OS, your IT workers could have fast access to their important data – providing that either your IT department or the OEM implements the add-in (see below). Possible scenario: Your staff can quickly glance at their Outlook e-mails or calendars without booting up the notebook. UEFI and its applications are on-screen within seconds.
Why the Slow Adoption?
It’s a simple matter of resource investment on the OEM side: UEFI is still a framework. Thus, to create unique pre-boot environments, add-ins, diagnostic tools, and specific hardware support, OEMs need to invest heavily into software development and tools.
Several companies are rolling out UEFI on a large scale, including Apple, IBM, and HP. This slide shows what HP has been doing in this business over the past two years:
It’s a major undertaking and many companies have held back until now.
You cannot upgrade your current PC fleet to UEFI. There is no feasible way to upgrade the BIOS chip on motherboards. From a technical point of view, it’s a completely different piece of software. Think of UEFI as an important aspect for future IT equipment decisions. Furthermore, only the 64-Bit versions of Windows 7, Windows Vista (SP1), and its bigger server editions (Windows Server 2008, Server 2008 R2) support the UEFI platform natively. 32-bit-platforms are only supported by UEFI mainboards with a special BIOS compatibility layer (CSM); this is something to look for if your infrastructure requires 32-bit operating systems.
If you think your IT infrastructure might benefit from UEFI, then your next step is to take it for a test drive. If your OEM doesn’t offer the UEFI tools you require, invest in your own pre-boot tools. For further information, I recommend reading UEFI forums whitepaper on evaluating the platform. A basic element of experimenting with UEFI is tianacore.
Predictions and Evaluations
The new “BIOS” is on the rise. The UEFI forum expects OEMs to finally adopt their new firmware on a large scale: In 2011, they predict, you will see more desktop clients equipped with UEFI than its predecessor. So what does this mean for you?
If the benefits we outlined in this article are important to you, then it’s time to evaluate UEFI sooner rather than later. You need to find a suitable hardware platform or partner that incorporates the UEFI software you need and start a small-scale test drive.
If you decide to hold off for now, then make no mistake: The transition is inevitable. UEFI will probably come with the next fleet of desktop PCs and notebooks you purchase.