Abstract: In accordance with the present invention, an apparatus includes a system bus having memory bank available signals. Coupled to the system bus are at least two memory modules, each having at least one memory bank. Each memory module includes a mechanism for associating each memory bank with one of the memory bank available signals. Further, each memory module includes logic for determining an availability status of each memory bank and for providing the associated memory bank busy signal with values reflecting the availability status of the memory bank. Additionally, at least two commander modules are coupled to the system bus and include logic, responsive to the memory bank available signals for preventing the commander module from gaining control of the system bus when the commander is attempting to access a memory bank determined to be unavailable. With such an arrangement, only commander modules seeking to access memory banks which are available will be allowed to gain control of the system bus.

Abstract: A commander module, coupled to a system bus including system bus control request signals and associated with one of the system bus control request signals, including means for determining whether control of the system bus is required and means for requesting control of the system bus, prior to determining whether such control is required, by asserting the associated system bus control request signal. A computer system including the system bus and at least two such commander modules coupled to the system bus and means for arbitrating for control of the system bus where the arbitrating means are coupled to and responsive to the system bus control request signals.

Abstract: 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.

Abstract: A data bus sequencer for use by nodes coupled to a system bus for associating data transactions and address transactions on the bus. The data bus sequencer includes means for tracking address and command transactions occurring on an address bus, the means for tracking producing a sequence number tag corresponding to each address and command transaction occurring on the address bus. Means for associating data transactions with address and command transactions stores the sequence number tags corresponding to address and command transactions for which data transactions are to be initiated by the node. Further included are means for tracking data transactions occurring on a data bus, means for comparing tracked data transactions to associated data transactions, and means for initiating data transactions on the data bus in response to the comparison.

Abstract: In accordance with the present invention, an apparatus includes a system bus having memory bank identification signals. Coupled to the system bus are at least two memory modules, each having at least one memory bank, and at least one commander module. The commander module contains decode logic which includes memory mapping registers associated with unique values to be driven on the memory bank identification signals. The memory banks contain compare logic including a virtual node identification register which stores a predetermined value to be compared with the value driven on the memory bank identification signals to determine if the memory bank is the target of the current transaction. Thus, memory banks need not decode the entire system bus address to determine if they are the target of the transaction which reduces the time required to complete a transaction with memory.

Abstract: A mechanism for ensuring the accurate and timely completion of atomic transactions by multiple nodes coupled to a memory via a common interconnect in a multiprocessor system includes a plurality of nodes coupled to a bus, the plurality of nodes including memory nodes, I/O nodes, and processor nodes. The memory nodes are each apportioned into a plurality of banks and together comprise the memory. Associated with each bank is a busy signal, indicating the availability of the bank of memory for transactions. A node may issue an atomic transaction to a block of memory data through the use of READ.sub.-- BANK.sub.-- LOCK and WRITE.sub.-- BANK.sub.-- UNLOCK instructions. The node executing the atomic transaction monitors the state of the busy signals of the banks, and when the bank is available, the node issues a READ.sub.-- BANK.sub.-- LOCK instruction, which sets the busy bit to indicate the unavailability of the bank. Upon the completion of the READ.sub.-- BANK.sub.-- LOCK instruction, the node issues a WRITE.

Abstract: A commander module including means for determining whether control of a system bus is required, means for requesting control of the system bus, prior to determining whether such control is required, and means, responsive to the determining means, for indicating that control of the system bus is required. A computer system including the system bus and at least two such commander modules coupled to the system bus and including means for arbitrating for control of the system bus including means for granting control of the system bus to one of the commander modules indicating that control of the system bus is required and having the highest arbitration priority among those commander modules also indicating that control of the system bus is required.

Abstract: Bus interfaces for nodes coupled to a system bus in a computer system, the system bus including an address bus and a separate data bus. System bus operations include address and command transactions and data transactions. Data transactions occur on the data bus separately and independently of the occurrence of address and command transactions on the address bus. A bus interface may include any of a commander address bus interface means for providing to an address bus address and command transactions, a responder address bus interface means for acknowledging receipt of address and command transactions via the address bus, a commander data bus interface means for controlling submission to the data bus of data transactions as a result of the occurrence of address and command transactions on the address bus, and a responder data bus interface means for transferring data on the data bus during a data transaction.

Abstract: A data bus sequencer for use by nodes coupled to a system bus for associating data transactions and address transactions on the bus. A mechanism for tracking address and command transactions occurring on the bus produces, for each address and command transaction occurring on the address bus, a corresponding sequence number tag. Those sequence number tags corresponding to address and command transactions for which data transactions are to be initiated by the node are stored by the data bus sequencer. The data bus sequencer further includes circuitry for counting the number of data transactions occurring on the data bus, comparing the counted number of data transactions to the stored sequence number tags and initiating data transactions on the data bus in response to the comparison.

Abstract: A synchronous computer system is described. The system is a multiprocessor system having a bus system clock and a processor clock for each processor. The system includes a synchronous computer system bus and a plurality of circuit modules coupled to the synchronous bus with at least two of the modules having at least one processor, with the processor modules having the at least one processor which runs asynchronously to each of the other processors while the processor modules are synchronous to the system bus. The system further includes clock generator means for providing a corresponding plurality of clock signals and a plurality of conductors coupled between said clock generating means and said plurality of modules.

Abstract: A memory system is provided which can adapt to being coupled to a bus capable of running at different clock speeds. The memory system is responsive to signals provided by a bus speed sensor for modifying the timing of row address strobe (RAS), column address strobe (CAS) and write enable (WE) signals. By modifying the timing of the RAS, CAS, and WE signals, the memory can be operated in systems capable of operating at a variety of bus speeds without suffering latency problems normally associated with changes in bus speed.

Abstract: In a computer system, digital signals are transmitted from an output register, propagated along a first signaling path, and received by an input register. The signaling path including an address buffer, a cache memory, a main memory, and an interconnect network. The effects of the intrinsic delays experienced by the digital signals are measured as a delay value relative to a reference clock signal propagated through a second signaling path duplicating the delays of the first signaling path. The delay value is used to selectively delay the digital signal to maintain a fixed relationship between the transmitted and received digital signals. Delay measuring and regulation is provided by driving the reference and digital signals through comparable tapped delay lines, the output taps of a measuring delay line controlling the output taps of a delaying line. Storage latches are provide to hold the measured delay value stable between successive samples.

Abstract: A load/store pipeline in a computer processor for loading data to registers and storing data from the registers has a cache memory within the pipeline for storing data. The pipeline includes buffers which support multiple outstanding read request misses. Data from out of the pipeline is obtained independently of the operation of the pipeline, this data corresponding to the request misses. The cache memory can then be filled with the requested for data. The provision of a cache memory within the pipeline, and the buffers for supporting the cache memory, speed up loading operations for the computer processor.

Abstract: A prediction logic device operating in conjunction with a vector processor to predict, before the completion of the translation of the virtual addresses of all of the data elements of a vector, the valid performance of all virtual-address to physical-address translations for the data elements of the vector. The prediction logic device asserts an MMOK signal to a scalar processor when it becomes known that no memory management fault and/or translation buffer miss will occur such that the scalar processor can resume vector instruction issue to the vector processor at the earliest possible time.

Abstract: A prediction logic device operating in conjunction with a vector processor to predict, before the completion of the translation of the virtual addresses of all of the data elements of a vector, the valid performance of all virtual-address to physical-address translations for the data elements of the vector. The prediction logic device asserts an MMOK signal to a scalar processor when it becomes known that no memory management fault and/or translation buffer miss will occur such that the scalar processor can resume vector instruction issue to the vector processor at the earliest possible time.

Abstract: A package for integrated circuit chips, or other electrical devices, provides mechanical shock and thermal protection for the chips, and in addition, protects the chips from electromagnetic interference and electrostatic discharge. The package includes a printed wiring board base for reception of one or more circuit chips, and a conductive heat sink and cover. The conductive heat sink, in conjunction with a reference plane in the wiring board base, acts as an EMI shield for the chips. The heat sink is covered with an insulating layer, on top of which, a conductive coating is placed. The conductive coating is electrically connected to the reference plane, and the two act to protect the chips from electrostatic discharges. Compliant pads support the chips, and a thermally conductive elastomer can be placed on top of each chip between the chips and the inner top surface of the heat sink. The chips are thereby held securely in position, and are thermally connected to the heat sink.

Abstract: A load/store pipeline in a computer processor for loading data to registers and storing data from the registers has a cache memory within the pipeline for storing data. The pipeline includes buffers which support multiple outstanding read request misses. Data from out of the pipeline is obtained independently of the operation of the pipeline, this data corresponding to the request misses. The cache memory can then be filled with the data that has been requested. The provision of a cache memory within the pipeline, and the buffers for supporting the cache memory, speed up loading operations for the computer processor.