Other Interfaces and Technologies: Contents at A Glance
Other Interfaces and Technologies: Contents at A Glance
Other Interfaces and Technologies: Contents at A Glance
Instant ON
Implementing instant ON
Device Bay
Implementing device bay
Further Study
978
979
Although this book presents many important PC technologies and bus schemes, there are
always a number of emerging interfacesboth software and hardware baseddesigned to streamline the performance and usability of the personal computer. Most new interfaces and technologies are designed around four key areas of the PC: improving graphics performance (AGP), improving power management (ACPI, APM, Instant ON, SMBus), simplifying connections and improving the performance of peripheral devices (Device Bay, I2O, USB), and improving the utilization of PCs (DMI, IrDA). Many of these emerging standards come from the PC industry leaders, such as Intel, Compaq, and Microsoft, but other standards are evolving from industry Special Interest Groups (SIGs), such as the Desktop Management Task Force (DTMF). Virtually all current PCs support one or more of these new standards to some degree, so as a technician, you should at least be familiar with the benefits and features that each one offers. This chapter presents an overview of these new interfaces and technologies, and offers some resources for more detailed study.
2
SYSTEM DATA AND TROUBLESHOOTING
puter into and out of system sleeping states. It also provides a general mechanism for any device to wake the computer, as needed. s Device power management ACPI tables describe motherboard devices, their power states, the power planes the devices are connected to, and controls for putting devices into different power states. This enables the OS to put various devices into low-power states, based on application usage. In other words, devices that are not being used by an application can be placed into low-power states, and devices that are being used can be kept active. s Processor power management While the OS is idle (but not sleeping), it will use commands described by the ACPI standard to put processors in low-power states. s Plug-and-Play ACPI specifies the techniques used to enumerate and configure motherboard devices. This information is arranged hierarchically, so when events, such as
980
docking and undocking, occur, the operating system has precise knowledge of which devices are affected by the event. System events ACPI provides a general event mechanism that can be used to keep track of thermal events, power-management events, docking, device insertion and removal, etc. This mechanism is very flexible in that it does not define specifically how events are routed to the core logic chipset. This allows individual device makers to incorporate their own responses to system events. Battery management For portable computers, battery-management policy moves from the APM BIOS to the ACPI operating system. The OS determines the low battery and battery warning points, and the OS also calculates the remaining battery capacity and remaining battery life. An ACPI-compatible battery device needs either a Smart Battery subsystem interface (SMbus), which is controlled by the OS directly through the embedded controller interface or a Control Method Battery (CMBatt) interface. Thermal management Because the ACPI OS controls the power states of devices and processors, ACPI also addresses system thermal management. It provides a simple, direct approach that allows manufacturers to define thermal zones, thermal indicators, and methods for cooling thermal zones. Embedded controller ACPI defines a standard hardware and software communications interface between an OS bus enumerator and an embedded controller. This allows any OS to provide a standard bus enumerator that can directly communicate with an embedded controller in the system, thus allowing other drivers within the system to communicate with and use the resources of system embedded controllers. This then enables an OEM to provide platform features that the OS and applications can use. System-management bus controller ACPI defines a standard hardware and software communications interface between an OS bus driver and an SMBus controller. This allows any OS to provide a standard bus driver that can directly communicate with SMBus devices in the system. This then enables an OEM to provide platform features that the OS and applications can use.
IMPLEMENTING ACPI
ACPI support is provided at the hardware and software level in a PC. At the hardware level, youll need a motherboard with an ACPI-compliant chipset, as well as ACPI compliant devices. For software, youll need an ACPI operating system (such as Windows 95) and perhaps an ACPI driver. Many of the newest PCs are now supporting ACPI. Keep in mind that using legacy devices might sometimes result in power-management or Plugand-Play problems under ACPI.
981
3D technology is that it requires intensive processing and lots of memory. This places a lot of strain on ordinary 3D video cards that use the PCI bustheir performance is often limited because of the massive amounts of data that must be passed between the video card and main memory. The AGP (Accelerated Graphics Port) developed by Intel uses a variation of the PCI bus slot to provide a high-speed data pathway between the 3D video card and main memory, and it allows the AGP video card to utilize main memory for graphics purposes. A prime example of AGPs potential is in the use of surface textures. Ordinarily, textures (the painted surfaces you see in so many 3D walk-through games, such as Quake II) are stored in the memory on a video card. This demands a lot of memory because texture maps are basically images, and textures are usually drawn in low-resolution to save space (which explains why so many scenes appear grainy when you get close to them). By moving the texture maps to main system RAM and accessing them across the AGP, it is possible to accelerate graphics performance because the 3D-rendering engine can focus on rendering, rather than patching in textures. The textures can also be made larger and at higher resolutions because there is typically much more main RAM available than graphics card memory. In actuality, other graphics-related data can also be moved to main memory through the AGP.
2
SYSTEM DATA AND TROUBLESHOOTING
latency.
s Demultiplexing of address and data on the bus, allowing almost 100% bus efficiency. s Timing for 133MHz data-transfer rates, allowing for real-data throughput in excess of
500MB/sec. The high-speed AGP is physically, logically, and electrically independent of the PCI bus. It is an additional expansion bus in the system (Fig. 30-1), which is intended for the exclusive use of visual display devicesall other I/O devices (such as drive controllers) will remain on the PCI bus. The add-in slot defined for AGP uses a new connector body (for electrical signaling reasons) which is not compatible with the PCI connector, so PCI and AGP boards are not mechanically interchangeable. The pinout for an AGP slot is listed in Table 30-1.
Although the AGP interface is based on the PCI bus, it was developed by Intel (independent of the PCI Special Interest Group) and has neither been reviewed nor endorsed by that group.
982
Processor
Graphics accelerator
AGP
Main chipset
System memory
TABLE 30-1 PINOUT FOR AN AGP SLOT PIN 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 SIDE B Spare 5.0 V 5.0 V USB+ GND INTB# CLK REQ# VCC (3.3) ST0 ST2 RBF# GND Spare SBA0 VCC (3.3) SBA2 SB_STB GND SIDE A 12 V Spare Reserved USBGND INTA# RST# GNT# VCC (3.3) ST1 Reserved PIPE# GND Spare SBA1 VCC (3.3) SBA3 Reserved GND
983
TABLE 30-1 PINOUT FOR AN AGP SLOT (CONTINUED) PIN 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 SIDE B GND SBA4 SBA6 Key Key Key Key AD31 AD29 VCC (3.3) AD27 AD25 GND AD_STB1 AD23 Vddq (3.3) AD21 AD19 GND AD17 C/BE2# Vddq (3.3) IRDY# GND VCC (3.3) DEVSEL# Vddq (3.3) PERR# GND SERR# C/BE1# Vddq (3.3) AD14 AD12 GND AD10 AD8 Vddq (3.3) AD_STB0 AD7 GND SIDE A GND SBA5 SBA7 Key Key Key Key AD30 AD28 VCC (3.3) AD26 AD24 GND Reserved C/BE3# Vddq (3.3) AD22 AD20 GND AD18 AD16 Vddq (3.3) FRAME# GND VCC (3.3) TRDY# STOP# Spare GND PAR AD15 Vddq (3.3) AD13 AD11 GND AD9 C/BE0# Vddq (3.3) Reserved AD6 GND
2
SYSTEM DATA AND TROUBLESHOOTING
984
TABLE 30-1 PINOUT FOR AN AGP SLOT (CONTINUED) PIN 61 44 62 63 64 65 66 SIDE B GND AD5 AD3 Vddq (3.3) AD1 SMB0 SIDE A GND AD4 AD2 Vddq (3.3) AD0 SMB1
IMPLEMENTING AGP
Implementing AGP on the desktop requires support from the latest hardware. Youll need a motherboard (often a Pentium II ATX or NLX motherboard) with a recent chipset that supports an AGP slot, as well as an AGP video adapter (e.g., one of the Matrox AGP video cards). With the suitable hardware available, youll need a video driver for Windows 95.
985
BIOS in a stand-alone configuration). The APM BIOS stand-alone power-management functions are enhanced once an APM driver establishes a connection with the APM BIOS. Once made, this connection establishes a protocol that allows the firmware to communicate power-management events to the APM driver, and to wait for APM driver acknowledgment, if necessary.
The operating system layer The APM driver has three primary power-management
functions: (1) passing calls and information between the application and APM BIOS layers, (2) arbitrating application power-management calls in a multitasking environment, and (3) identifying power-saving opportunities not apparent at the application or BIOS layer. An APM connection must be established between the APM BIOS and an APM driver for optimum system power management.
TABLE 30-2 APM SYSTEM CONDITIONS APM STATE Full On SYSTEM CONDITIONS System is working. System is not power managed. All devices are on. System is working. System is power managed. The CPU clock is slowed or stopped as needed. Devices are power managed as needed. System might not be working. System is in a low-power state with some power savings. Most devices are in a low-power mode. The CPU clock is slowed or stopped. Operational parameters are retained. System returns quickly to the APM-enabled state. The Resume timer event must return the system to the APM-enabled state. User might be required to return the system to the APM-enabled state. The operating system is notified after the system shifts to the Enabled state. Prior operation resumes after returning to the APM-enabled state. Interrupts must be processed normally (this might require waking up the CPU temporarily if it was stopped, but the CPU might be stopped again by the APM driver). System is not working. System is in a low-power state with maximum power savings. Most power-managed devices are not powered. The CPU clock is stopped. The CPU core is in its minimum-power state. Operational parameters are saved to be restored later when resuming. System takes a relatively long time to return to the APM-enabled state. Wakeup events can return the system to the APM-enabled state. The Resume timer event must be one of the wakeup events. The OS is notified after the system shifts to the Enabled state. Prior operation resumes after returning to the APM-enabled state. System is not working. The power supply is off. Operational parameters are not saved. System resets and initializes when shifting to the Full-on state.
2
SYSTEM DATA AND TROUBLESHOOTING
APM Enabled
APM Standby
APM Suspend
Off
986
Applications
APM-aware application
APM-aware application
OS dependent Operating system APM driver APM interface BIOS APM BIOS OS dependent
Add-in device
Add-in device
By regularly polling the APM BIOS, the APM driver will determine whether the APM BIOS wants a power-saving state to occur. In the case of a Standby or Suspend request, the APM driver is expected to do the appropriate processing to prepare for the state change, then call the APM BIOS to actually execute the power-state change in hardware. Because only one APM driver can exist in the system (and there might be many APM applications), the APM driver must provide an interface between the APM BIOS and APM-aware applications. This interface should pass the applications APM requests to the APM BIOS, and send any APM BIOS-generated events back to the applications. Each APM driver should specify a power-management application-to-OS interface.
The application layer APM-aware applications assist in power management by pro-
viding information that only the application is in a position to know (or easily determine). Similarly, device drivers for add-in devices that are not under direct control of the BIOS (such as PCMCIA cards or ISA video adapters) might be able to manage power on their device, given suitable information about the overall system state, or by monitoring usage of their own hardware. Applications and device drivers are not required to be APM-aware, but they can greatly increase APM effectivenessparticularly on less-sophisticated operating systems. Under DOS, the application or device driver is often in the best position to know when the application or an add-in device is idle and awaiting further activity. APM-aware applications register with the APM driver using an OS-dependent mechanism. The APM driver notifies registered applications and device drivers when system power-management events occur, and the applications and device drivers then take suitable action. For example, when an APM-aware device driver learns from the APM driver that the system will be suspended, it saves device information to be restored when the system resumes. When an APM-aware device driver is notified that the system has resumed, it restores the add-in device to its previous operating state.
987
IMPLEMENTING APM
APM requires four elements to function properly: an APM BIOS, an APM operating system (i.e., Windows 95), power-manageable devices, and APM device drivers to operate those devices. Most Pentium and Pentium MMX PCs support APM, although the very newest PCs (e.g., Pentium II systems) are implementing ACPI support instead of APM. Now in version 1.2, it is unlikely that the APM specification will continue to develop as a stand-alone technology. The reason for this is that APM techniques are known to interfere with some Plug-and-Play devices (especially PC Card swapping and mobile PC docking and undocking operations). Rather than continue developing APM, the role of APM has been expanded and combined with PnP operations in the ACPI standard (see the ACPI section in this chapter).
Device Bay
Adding or replacing a drive on a PC continues to be a fairly labor-intensive process. Even with the benefits of automated device detection and driver installation offered by Windows 95, it is still necessary for users or technicians to physically open the system and manually install or replace any given device. To simplify system assembly, maintenance, and expansion, designers have introduced the Device Bayan industry specification for interchangeable peripheral devices (such as hard-disk drives, modems, network adapters, CD-ROM drives, DVD-ROM drives, and a variety of other electronic devices). When fully implemented, you should simply insert a peripheral (e.g., a DVD-ROM) into the PC without opening, rebooting, or even powering down the PC. The system would then recognize the new or replaced device, and automatically reconfigure the system to accommodate it. Device Bay is expected to use a combination of USB port for hot-swappable, mediumspeed serial data, and IEEE 1394 Fire Wire for high-speed serial data. In addition, the Device Bay will accommodate staged power management (compatible with ACPI).
2
SYSTEM DATA AND TROUBLESHOOTING
988
or Computer City, and youll see a vast array of devices. There are literally dozens of options to choose from in everything from cases to CD-ROM drives and sound cards. You can also see this range of products in the myriad of PCs now in the market. Unfortunately, the problem with all of this diversity is that it is very difficult to manage from the technical standpoint. In other words, its hard to find out about whats inside any given PC. Diagnostics can identify some parts, but diagnostics are not always reliable, and they are obsolete soon after new products arrive. Diagnostics also dont provide useful information about when products were installed or upgraded, or what versions of drivers are in use. This often means that a technician must work in the dark. Some corporate environments might have hundreds (perhaps thousands) of computers that must be networked and supported, so you can imagine how daunting the task of equipment management can be for managers and technicians alike. The Desktop Management Task Force (DTMF) has developed the Desktop Management Interface (DMI) as a solution to this information gap. The DMI is a software standard for describing and accessing information about all types of PCs and PC components. The DMI standard provides a common resource for tech support, IT managers, and individual users to access information about all aspects of a given PC, including processor type, installation date, attached printers and other peripherals, power sources, and maintenance history. The DMI provides support for describing the more than 80,000 PC products in the marketplace today, allowing more cost-effective (and less crisis-driven) PC management and support. Current versions of DMI also allow technicians to troubleshoot remote PCs right over a network from their own desktop PC.
IMPLEMENTING DMI
Now at version 2.0, the DMI is a software standard requires a service layer of software under the operating system, as well as DMI-enabled products (devices with .MIF files already prepared). DMI will work under Windows 95, and many major system manufacturers (such as Compaq, Dell, HP, IBM, NEC, and Sun) are now producing DMI-compliant systems. There are now about 200 to 300 DMI-enabled devices. As more devices are instrumented for DMI, it is very probable that DMI-enabled systems will become more popular. For an individual end-user, DMI has little direct impact on your system. It does not improve performance in any wayit simply provides a standard suite of information about whats inside the machine. Such information has little value for end users, but might be handy when troubleshooting or upgrading a system. As a consequence, you will probably encounter DMI-enabled systems, but it is unlikely that you would upgrade a system to
989
support DMI. However, such standards as DMI might eventually make remote troubleshooting services feasible for end-user systems.
2
SYSTEM DATA AND TROUBLESHOOTING
I2O IN OPERATION
I2O defines a standard architecture for intelligent I/Oan approach to I/O in which lowlevel interrupts are off-loaded from the CPU to I/O sub-processors (IOPs) designed specifically to handle I/O processing. There is support for message-passing between multiple independent I/O processors, so the I2O architecture relieves the host of interrupt-intensive I/O tasks, greatly improving I/O performance in high-bandwidth applications, such as networked video, groupware, and client/server processing. I2O also provides support for single-processor, multiprocessor, and clustered systems. The I2O specification also defines a universal driver approach for creating drivers that are portable across multiple operating systems and host PC platforms. With the proliferation of Network Operating Systems (NOSs), most notably NetWare 4, Windows NT
990
Server, and UnixWare, the number of drivers that must be written, tested, integrated and supported has escalated (one for every unique combination of OS and device). Using the universal driver approach, I2O significantly decreases the number of drivers required. OS vendors write a single I2O-ready driver for each class of device (such as disk adapter), known as the OS Services Module (OSM) and device manufacturers write a single I2Oready driver for each device, known as the Hardware Device Module (HDM), which will work for any OS that supports I2O. This two-driver model ensures that the OS does not have to be aware of every deviceonly classes of devices, and each particular device will offer its own I2O driver, which will interface to the OSs I2O device class driver.
IMPLEMENTING I2O
Although I2O was developed by the I2O SIG in 1996, the actual implementation of I2O on desktop systems has been slowlargely because the I2O API needed for operating system support has not yet appeared for Windows 95 or Windows NT 4.0 (although it might appear in Windows 98 or Windows NT 5.0), and I2O-compliant devices are slow in appearing. Also, although chipsets are now available to support I2O, few desktop-class motherboards are incorporating the I/O sub-processors and PCI bridge needed to bring intelligent I/O to a PCI bus. It is possible to provide the I/O sub-processor(s) on a PCI expansion card (making it possible to add I2O support by adding the expansion card), but PCI slots are often too scarce for this tactic. Finally, few existing devices in the field actually support I2O at this time (the drivers are also limited). The I2O SIG expects general implementation of I2O by mid-1998.
Instant ON
One of the problems with power conservation has traditionally been connectivitythe PC cannot respond to the outside world while in a power-saving mode, and any network or dial-up connection already in service is usually severed. Another problem has often been that power recovery after being in a power-saving state is often a slow and convoluted process. Instant ON (a part of the APM and broad ACPI standards) resolves these issues by allowing fast recovery of the system when needed, and support for remote waking by external events (such as incoming faxes) or scheduled events (such as backups or defragmenting the hard drive). This improved functionality allows you to keep the PC turned on at all times and be responsive to real-world events, yet provide excellent power conservation. Instant ON provides three key features:
s Schedule automated tasks With Instant ON, you no longer need to be present for your
PC to be productive. Programs can be run whether you are at the machine or not (and whether the machine is in use or not). Schedule disk maintenance, phone calls, and other routine tasks whenever they are most convenient. Leave your Windows 95 system powered up at all times, and the Instant ON scheduler will run the tasks youve specified. You can set up applications to run at particular times or on particular days. You can run programs repeatedly (e.g., once each week on Mondays or once a month on the first of the month). You can schedule programs to run several times a day if
991
needed. They can also be scheduled to run after another application has finished a scheduled task. Instant ON also switches your PC to low-power mode to save energy and wear. If the machine is in low power, Instant ON powers it up in time to run a scheduled task. s Unattended operation Automating tasks also makes it possible for a computer to run in an unattended mode. For instance, you do not need to be present for your fax/modem software to receive an incoming fax, so it can run without you. The disk utilities supplied with the Microsoft Plus Pack are also designed to run using settings determined in advance. The program should also be able to stop without requiring confirmation or other input from you.
Applications that are interactive and assume that the user is present when they are running should not be run unattended. If the program pops up dialog boxes, asks for confirmation, or needs your input while it runs, do not schedule it to run when no one is at the computer. s Energy efficient, yet available With Instant ON, you never need to turn off the PC, so
2
SYSTEM DATA AND TROUBLESHOOTING
you dont have to wait for it to boot. Instant ON allows a power-managed PC to enter a low-power state when application and user activity cease, or when you choose to switch the PC to its low-power mode from the Instant ON scheduler. The system comes out of low-power mode when a scheduled task is to be run or when activity resumes (such as when the mouse is moved). The low-power state provides energy savings while allowing the system to return immediately to full power whenever necessary.
IMPLEMENTING INSTANT ON
Instant ON is actually a software application developed by Intel to enhance PCs with APM or ACPI power-conservation schemes in place. To use Instant ON, youll need the Instant ON applet (available from http://www.intel.com), a PC platform that supports APM or ACPI power management, and power-managed devices (such as a modem).
992
s Sending faxes/e-mail from a notebook PC through an IrDA-equipped telephone. s Banking using a mobile PC and IrDA-equipped ATM. s Accessing a network with a mobile PC using an IrDA-equipped node.
IRDA STANDARDS
In September 1993, IrDA outlined the basis for the IrDA SIR Data Link Standards. In June 1994, IrDA published the IrDA standards, which includes Serial Infrared (SIR) Link specification, Link Access Protocol (IrLAP) specification, and Link Management Protocol (IrLMP) specification. IrDA released extensions to SIR standard (including 4Mb/s) in October 1995. The IrDA standard specification has been expanded to include high-speed extensions from 1.152 Mb/s and 4.0 Mb/s. This extension will require an add-in card to retrofit existing PCs with high-speed IR and synchronous communications controllers (or equivalent).
LIMITATIONS TO IRDA
Although the IR serial port offers some serious advantages to PC users, IR technology has some limitations that you should keep in mind before using or upgrading to an IrDA port:
s IR working range is limited from 1 m to about 3 m. s You need a direct line of site between two IR ports. s The two IR ports must be directly facing one another (an angle of no more than 30
degrees).
s IrDA drivers and software further complicate an already crowded Windows 95 platform.
IMPLEMENTING IRDA
IrDA ports are now being incorporated into many new mobile computers and peripherals (especially printers). If you already have IrDA-enabled devices, you just need IrDA software and drivers for Windows 95. If you want to upgrade existing desktop systems, mobile systems, or peripherals, you can purchase and install IrDA adapters for existing serial ports (including the software and drivers needed to operate the IrDA adapters).
993
Applications and device drivers Power management API Operating system Smart battery BIOS spec APM BIOS SMB BIOS
Charger controller
SMB host
2
SYSTEM DATA AND TROUBLESHOOTING
SMBus host over a two-wire serial interface. SMBus BIOS provides the low-level BIOS support needed to operate the SMBus host, and an SMBus API provides the high-level operating-system support and dialog boxes used to configure and operate the SMBus.
Alarms The battery is about to run out. Control/Status/Error To control the batterys operating mode. Predictive functions Indicating how long will the battery will supply power at a given rate. Measured data Indicting how much current the battery is supplying.
994
s s s s
Battery state of charge The fuel gauge information. Charging information Indicating how to properly charge the battery. Battery characteristics Outlining the design capacity and cell chemistry of the battery. Manufacturing data Providing the manufacturers name and serial number.
IMPLEMENTING SMBUS
Because the SMBus was designed primarily for mobile PCs, you will probably encounter SMBus and Smart Battery configurations, but you cannot add SMBus support to an existing system. The most important impact youll have on smart battery operations is the proper configuration of Smart Battery software.
USB OPERATIONS
USB has two data rates: a 12Mbps rate for devices requiring increased bandwidth, and a 1.5Mbps rate for lower-speed devices, such as joysticks and game pads. USB uses a tiered star topology, which means that some USB devices (called USB hubs) can serve as connection ports for other USB peripherals. Only one device needs to be plugged into the PC. Other devices can then be plugged into the hub. USB hubs can be embedded in such key devices as monitors, printers and keyboards. Stand-alone hubs can also be made available. Hubs feature an upstream connection (pointed toward the PC), as well as multiple downstream ports to allow the connection of additional peripheral devices. Up to 127 USB devices can be connected together in this way. Industry pundits often refer to this as plug and play outside of the box. USB host controllers (which are available as part of several Intel PCI chip sets) manage and control the driver software and data flow required by each peripheral connected to the bus. Users dont need to take any specific configuration action because all the configuration steps happen automaticallythe USB host controller even allocates electrical power to the USB devices. USB hubs and host controllers can detect attachments and detachments of peripherals occurring downstream and supply appropriate levels of power to downstream devices, as needed. Figure 30-4 illustrates a typical USB connector and pinout.
FURTHER STUDY
995
USB port 0 Pin1 Pin2 Pin3 Pin4 Power (fused) Signal# Signal Ground
2
SYSTEM DATA AND TROUBLESHOOTING
IMPLEMENTING USB
Most new systems are equipped with one or two USB ports (usually located in the area of COM or LPT ports). In this case, it is simply a matter of attaching a USB hub (such as a USB keyboard), then attaching USB devices to the hub. For systems without USB, youll need a motherboard upgrade, which contains a USB-compliant chipset and port(s). Most current ATX and NLX motherboards will sport a USB port. Once the new motherboard is in place, USB devices can be attached.
Further Study
That concludes the material for Chapter 30. Be sure to review the glossary and chapter questions on the accompanying CD. If you have access to the Internet, take a look at some of these resources: ACPI: http://www.teleport.com/~acpi/ APM: http://www.intel.com/IAL/powermgm/ AGP: http://www.agpforum.org/ ATX: http://www.teleport.com/~atx Device Bay: http://www.device-bay.org DMI: http://www.dmtf.org/
996
I2O: http://www.i2osig.org/ IrDA: http://www.irda.org/ Smart Battery and SMBus: http://www.sbs-forum.org/ Instant ON Scheduler: http://developer.intel.com/ial/inston/install.htm USB: http://www.usb.org/