Compaq Computer Corporation, Inc.

Digital Equipment Corporation

Stratus Technologies Ireland Ltd.

Craig Durand Keefer

Samuel Hammond Duncan

Paul Michael Guglielmi

Ricky C Hetherington

Glenn Arthur Herdeg

Maurice Bennet Steinman

Thomas Adam McLaughlin

Simon C Steely, Jr

Madhumitra Sharma

Stephen R VanDoren

Andrew Alden

Michael Egan

John M Chaves

Chester W Pawlowski

David W Davis

Christopher D Cotton

A fault tolerant computer system and method are disclosed. The system may include a plurality of CPU nodes, each including: a processor and a memory; at least two IO domains, wherein at least one of the IO domains is designated an active IO domain performing communication functions for the active CPU nodes; and a switching fabric connecting each CPU node to each IO domain. One CPU node is designated a standby CPU node and the remainder are designated as active CPU nodes. If a failure, a beginning of a failure, or a predicted failure occurs in an active node, the state and memory of the active CPU node are transferred to the standby CPU node which becomes the new active CPU node. If a failure occurs in an active IO domain, the communication functions performed by the failing active IO domain are transferred to the other IO domain.

High reliability fault tolerant computer architecture

A multiprocessor having improved bus efficiency is shown to include a number of processing units and a memory coupled to a system bus. Also coupled to the system bus are at least one I/O bridge systems. A method for improving partial cache line writes from I/O devices to the central processing units incorporates cache coherency protocol and an enhanced invalidation scheme to ensure atomicity, which minimizes the bus utilization. In addition, a method for allowing peer-to-peer communication between I/O devices coupled to the system bus via different I/O bridges includes a command and address space configuration that allows for communication without the involvement of any central processing device. Interrupt performance is improved through the storage of an interrupt data structure in main memory. The I/O bridges maintain the data structure, and when the CPU is available the interrupts can be accessed by a fast memory read; thereby reducing the requirement of I/O reads for interrupt handling.

Performance optimization and system bus duty cycle reduction by I/O bridge partial cache line write

An architecture and coherency protocol for use in a large SMP computer system includes a hierarchical switch structure which allows for a number of multi-processor nodes to be coupled to the switch to operate at an optimum performance. Within each multi-processor node, a simultaneous buffering system is provided that allows all of the processors of the multi-processor node to operate at peak performance. A memory is shared among the nodes, with a portion of the memory resident at each of the multi-processor nodes. Each of the multi-processor nodes includes a number of elements for maintaining memory coherency, including a victim cache, a directory and a transaction tracking table. The victim cache allows for selective updates of victim data destined for memory stored at a remote multi-processing node, thereby improving the overall performance of memory. Memory performance is additionally improved by including, at each memory, a delayed write buffer which is used in conjunction with the directory to identify victims that are to be written to memory. An arb bus coupled to the output of the directory of each node provides a central ordering point for all messages that are transferred through the SMP. The messages comprise a number of transactions, and each transaction is assigned to a number of different virtual channels, depending upon the processing stage of the message. The use of virtual channels thus helps to maintain data coherency by providing a straightforward method for maintaining system order. Using the virtual channels and the directory structure, cache coherency problems that would previously result in deadlock may be avoided.

High-performance non-blocking switch with multiple channel ordering constraints

A multiprocessor having improved bus efficiency is shown to include a number of processing units and a memory coupled to a system bus. Also coupled to the system bus are at least one I/O bridge systems. A method for improving partial cache line writes from I/O devices to the central processing units incorporates cache coherency protocol and an enhanced invalidation scheme to ensure atomicity which minimizing the bus utilization. In addition, a method for allowing peer-to-peer communication between I/O devices coupled to the system bus via different I/O bridges includes a command and address space configuration that allows for communication without the involvement of any central processing device. Interrupt performance is improved through the storage of an interrupt data structure in main memory. The I/O bridges maintain the data structure, and when the CPU is available the interrupts can be accessed by a fast memory read; thereby reducing the requirement of I/O reads for interrupt handling.

Performance optimization and system bus duty cycle reduction by I/O

Method and apparatus for providing DMA transfers between devices coupled

Method and apparatus providing DMA transfers between devices coupled to

Growing community of inventors

Nashua, NH, United States of America

Craig Durand Keefer