What's UP With [I.sub.2]O SIG?Work started on the initial [I.sub.2]O specification in 1995, the product of a strategic discussion between engineers at Intel and Netframe Systems. The majority of the technical work on [I.sub.2]O was done in 1996 and 1997, when several plugfests were held where developers from the U.S., Europe, and Asia met to test hardware and software prototypes. Revision 1.5 of the specification was released in 1997, made public in 1998, and over 500 developers registered on the [I.sub.2]O SIG web site to receive a copy. The [I.sub.2]O specification defines a "grand architecture" for servers in which one or more I/O Processors (IOPs) are provided, which are separate from the main server CPU CPU in full central processing unit Principal component of a digital computer, composed of a control unit, an instruction-decoding unit, and an arithmetic-logic unit. . These IOPs control interfaces, which are mounted on the motherboard or implemented as plug-in cards. A queued message-based protocol for CPU to IOP IOP intraocular pressure. IOP Intraocular pressure, see there communication is included, with a common packet format, flexible scatter/gather facilities, and dynamic configuration support. Specific message classes are defined to communicate with abstractions of block storage devices, SCSI SCSI in full Small Computer System Interface Once common standard for connecting peripheral devices (disks, modems, printers, etc.) to small and medium-sized computers. SCSI has given way to faster standards, such as Firewire and USB. peripherals, LAN (Local Area Network) A communications network that serves users within a confined geographical area. The "clients" are the user's workstations typically running Windows, although Mac and Linux clients are also used. interfaces, etc., and between peer IOPs. As a result of experience with products based on Rev 1.5, further development work was conducted in 1998, which led to Revision 2.0 of the specification being announced in May 1999 and which will shortly be released to developers. New features included in 2.0 are replacement scatter/gather list formats, support for 64-bit physical addressing and message buffering in host memory, and a message class for communication with hot-plug controllers. The complete [I.sub.2]O "grand architecture" is not found in many servers. However, the protocol, abstraction scheme, and message classes defined to support the architecture have been successfully used in a number of shipping products. The majority of the shipping [I.sub.2]O products use PCI (1) (Payment Card Industry) See PCI DSS. (2) (Peripheral Component Interconnect) The most widely used I/O bus (peripheral bus). as the transport mechanism for [I.sub.2]O messages. It should be noted, however, that the [I.sub.2]O architecture does not require the use of PCI, or even shared memory (1) Using part of main memory to support a low-cost display circuit that does not have its own memory. See shared video memory. (2) The common memory in a symmetric multiprocessing system that is available to all CPUs. See SMP. 1. , between the IOP and CPU. Hence, the [I.sub.2]O API (Application Programming Interface) A language and message format used by an application program to communicate with the operating system or some other control program such as a database management system (DBMS) or communications protocol. is also finding considerable use in the "switched fabric" interconnection schemes currently under development. Roger Cummings is the director of strategic relationships at DPT (Maitland, FL). www.dpt.com www.i2sig.org |
|
||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion