This International Standard provides a set of tools with which to implement a Futurebus+ architecture with performance and cost scalability over time, for multiple generations of single- and multiple-bus multiprocessor systems. Although this specification is principally intended for 64-bit address and data operation, a fully compatible 32-bit subset is provided, along with compatible extensions to support 128- and 256-bit data highways. Allocation of bus bandwidth to competing modules is provided by either a fast centralized arbiter, or a fully distributed, one or two pass, parallel contention arbiter. Bus allocation rules are provided to suit the needs of both real-time (priority based) and fairness (equal opportunity access based) configurations. Transmission of data over the multiplexed address/data highway is governed by one of two intercompatible transmission methods: a) a technology-independent, compelled-protocol, supporting broadcast, broad call, and transfer intervention (the minimum requirement for all Futurebus+ systems), and b) a configurable transfer-rate, source-synchronized protocol supporting only block transfers and source-synchronized broadcast for systems requiring the highest possible performance. Futurebus+ takes its name from its goal of being capable of the highest possible transfer rate consistent with the technology available at the time modules are designed, while ensuring compatibility with all modules designed to this standard both before and after. The plus sign (+) refers to the extensible nature of the specification, and the hooks provided to allow furthere evolution to meet unanticipated needs of specific application architectures. It is intended that this International Standard be used as a key component of an approved IEEE Futurebus+ profile.