Abstract: Methods for improving processor performance. Specifically, by reducing some of the latency cycles within a host controller, request processing speed can be improved. One technique for improving processing speed involves initiating a deferred reply transaction before the data is available from a memory controller. A second technique involves anticipating the need to transition from a block next request (BNR) state to a bus priority request (BPRI) state, thereby eliminating the need to wait for a request check to determine if the BPRI state must be implemented.

Abstract: An internal bus structure for a multi-processor-bus system. More specifically, an internal bus protocol/structure is described. The internal bus structure includes unidirectional, point-to-point connections between control modules. The individual buses carry unique transactions corresponding to a request. Each transaction includes an identification tag. The present protocol provides for efficient communication between processors, peripheral devices, memory and coherency modules. The present protocol and design scheme is generic in that the techniques are scalable and re-usable.

Abstract: A computer system includes an adaptive memory arbiter for prioritizing memory access requests, including a self-adjusting, programmable request-priority ranking system. The memory arbiter adapts during every arbitration cycle, reducing the priority of any request which wins memory arbitration. Thus, a memory request initially holding a low priority ranking may gradually advance in priority until that request wins memory arbitration. Such a scheme prevents lower-priority devices from becoming “memory-starved.” Because some types of memory requests (such as refresh requests and memory reads) inherently require faster memory access than other requests (such as memory writes), the adaptive memory arbiter additionally integrates a nonadjustable priority structure into the adaptive ranking system which guarantees faster service to the most urgent requests.

Abstract: A computer system includes a plurality of memory modules that contain semiconductor memory, such as DIMMs. The system includes a host/data controller that utilizes an XOR engine to store data and parity information in a striped fashion on the plurality of memory modules to create a redundant array of industry standard DIMMs (RAID). The host/data controller also interleaves data on a plurality of channels associated with each of the plurality of memory modules.

Abstract: A method of replacing a memory module in a computer system. Specifically, a method for replacing a memory module in a segment of a redundant memory system, without powering-down the memory system.

Abstract: A technique for selecting events associated with a hot-plug operation. More specifically, a programmable configuration register may be used to provide a mechanism for periodically scheduling requests associated with a hot-plug operation, such as initialization, rebuild, and verify requests. An arbiter is provided to facilitate an ordered access to a memory system. A user can select a periodic interval such that hot-plug requests are periodically executed during the execution of normal requests through the arbiter. The user-selectable interval may be dependent on the specific application of the system and the importance of operating in a redundant mode.

Abstract: A computer is provided having a bus interface unit coupled between a processor bus, a peripheral bus, and a memory bus. The bus interface unit includes a processor controller linked to the processor bus for controlling the transfer of cycles from the processor to the peripheral bus and memory bus. Those cycles are initially forwarded as a request, whereby the processor controller includes a memory request queue separate from a peripheral request queue. Requests from the memory and peripheral request queues can be de-queued concurrently to the memory and peripheral buses. This enhances throughput of read and write requests; however, proper ordering of data returned as a result of read requests and data transferred as a result of write requests must be ensured. An in-order queue is also present in the processor controller which records the order in which the requests are dispatched to the peripheral and memory buses from the peripheral and memory request queues.

Abstract: A method of adding memory capacity to a computer system. The computer system comprises a redundant memory system including a plurality of memory cartridges. By powering-down a memory cartridge, adding an additional memory module to the memory cartridge, and powering-up the memory cartridge for each memory cartridge in the system, the system can transition from a redundant mode of operation to a non-redundant mode of operation for each power-down, thus allowing the computer system to remain functional during the addition of the memory module. Alternatively, memory cartridges with higher memory capacity than those currently present in the computer system can be used to replace existing memory cartridges in the computer system, using the same techniques.

Abstract: The control logic for a hot-pluggable memory cartridge for use in a redundant memory system. To implement a hot-pluggable memory cartridge in a redundant memory system, control logic to control the sequence of events for powering-up and powering-down a memory cartridge is provided.

Abstract: A system and technique for correcting data errors in a memory device. More specifically, data errors in a memory device are corrected by scrubbing the corrupted memory device. Generally, a host controller delivers a READ command to a memory controller. The memory controller receives the request and retrieves the data from a memory sub-system. The data is delivered to the host controller. If an error is detected, a scrub command is induced through the memory controller to rewrite the corrected data through the memory sub-system. Once a scrub command is induced, an arbiter schedules the scrub in the queue. Because a significant amount of time can occur before initial read in the scrub write back to the memory, an additional controller may be used to compare all subsequent READ and WRITE commands to those scrubs scheduled in the queue.

Abstract: A system and technique for detecting data errors in a memory device. More specifically, data errors in a memory device are detected by initiating an internal READ command or verify operation from a set of logic which is internal to the memory system in which the memory devices reside. Rather than relying on a READ command to be issued from an external device, via a host controller, the verify logic initiates verify routine in response to an event such as an operator instruction, hot-plug operation, or a periodic schedule. By implementing the verify operation, the system does not rely on external READ commands to verify data integrity. The verify routine may rely on typical ECC error logging mechanisms and may be used in a RAID memory architecture. Further, the verify routine may be used in conjunction with other error logging and correction logic, as well as scrubbing logic.

Abstract: A computer system includes a plurality of memory modules that contain semiconductor memory, such as DIMMs. The system includes a host/data controller that utilizes an XOR engine to store data and parity information in a striped fashion on the plurality of memory modules to create a redundant array of industry standard DIMMs (RAID). The system supports DIMMs having X4 and X8 configurations. The system also transitions between various states, including a redundant state and a non-redundant state, to facilitate “hot-plug” capabilities utilizing its removable memory cartridges.

Abstract: A computer system includes an adaptive memory arbiter for prioritizing memory access requests, including a self-adjusting, programmable request-priority ranking system. The memory arbiter adapts during every arbitration cycle, reducing the priority of any request which wins memory arbitration. Thus, a memory request initially holding a low priority ranking may gradually advance in priority until that request wins memory arbitration. Such a scheme prevents lower-priority devices from becoming “memory-starved.” Because some types of memory requests (such as refresh requests and memory reads) inherently require faster memory access than other requests (such as memory writes), the adaptive memory arbiter additionally integrates a nonadjustable priority structure into the adaptive ranking system which guarantees faster service to the most urgent requests.

Abstract: A computer system includes a CPU and a memory device coupled by a bridge logic unit. CPU to memory write requests (including the data to be written) are temporarily stored in a queue in the bridge logic unit. The bridge logic unit preferably begins a write cycle to the memory device before all of the write data has been stored in the queue and available to the memory device. By beginning the memory cycle as early as possible, the total amount of time required to store all of the write data in the queue and then de-queue the data from the queue is reduced. Consequently, many CPU to memory write transactions are performed more efficiently and generally with less latency than previously possible.

Abstract: A computer is provided having a bus interface unit coupled between a CPU bus, a peripheral bus (i.e., PCI bus and/or graphics bus), and a memory bus. The bus interface unit includes controllers linked to the respective buses, and a plurality of queues placed within address and data paths between the various controllers. The peripheral bus controller can decode a write cycle to memory, and the processor controller can thereafter request and be granted ownership of the CPU local bus. The address of the write cycle can then be snooped to determine if valid data exists within the CPU cache storage locations. If so, a writeback operation can occur. Ownership of the CPU bus is maintained by the bus interface unit during the snooping operation, as well as during writeback and the request of the memory bus by the peripheral-derived write cycle. It is not until ownership of the memory bus is granted by the memory arbiter that mastership is terminated by the bus interface unit.

Abstract: A computer system, bus interface unit, and method are provided to allocate requests to a shared bus within the computer system. The bus interface unit includes an arbiter which employs a multi-level, round-robin arbitration protocol. Configuration registers are programmed during boot-up of the computer system by assigning a subset of peripheral devices, bus agents, requesters, or bus masters to either a high priority ring or a low priority ring, if two levels of arbitration are used. The status of a low priority device can be elevated to equal priority with a high priority device by assigning the low priority device to a high priority port within the high priority ring if certain circumstances occur. Namely, if data transfers to or from the low priority device are terminated, then the low priority device will be promoted to a high priority device so that it need not wait until after the all high priority device requests have been polled.

Abstract: A computer system includes an adaptive memory arbiter for prioritizing memory access requests, including a self-adjusting, programmable request-priority ranking system. The memory arbiter adapts during every arbitration cycle, reducing the priority of any request which wins memory arbitration. Thus, a memory request initially holding a low priority ranking may gradually advance in priority until that request wins memory arbitration. Such a scheme prevents lower-priority devices from becoming “memory-starved.” Because some types of memory requests (such as refresh requests and memory reads) inherently require faster memory access than other requests (such as memory writes), the adaptive memory arbiter additionally integrates a nonadjustable priority structure into the adaptive ranking system which guarantees faster service to the most urgent requests.

Abstract: A computer system that includes a CPU, a memory and a memory controller for controlling access to the memory. The memory controller generally includes arbitration logic for deciding which memory request among one or more pending requests should win arbitration. When a request wins arbitration, the arbitration logic asserts a “won” signal corresponding to that memory request. The memory controller also includes synchronizing logic to synchronize memory requests, corresponding to a first group of requests, that win arbitration to a clock signal and an arbitration enable signal. The synchronizing logic includes an AND gate and a latch for synchronizing the won signals. The memory controller also asynchronously arbitrates a second group of memory requests by asserting a won signal associated with the second group requests that is not synchronized to the clock signal.

Abstract: A computer system includes a CPU, a memory device, two expansion buses, and a bridge logic unit coupling together the CPU, the memory device and the expansion buses. The CPU couples to the bridge logic unit via a CPU bus and the memory device couples to the bridge logic unit via a memory bus. The bridge logic unit generally routes bus cycle requests from one of the four buses to another of the buses while concurrently routing bus cycle requests to another pair of buses. The bridge logic unit preferably includes four interfaces, one each to the CPU, memory device and the two expansion buses. Each pair of interfaces are coupled by at least one queue; write requests are stored (or “posted”) in write queues and read data are stored in read queues.

Abstract: A computer is provided having a bus interface unit coupled between a CPU bus, a peripheral bus, and a memory bus. The bus interface unit includes a processor controller linked to the CPU bus for controlling the transfer of cycles from the CPU to the peripheral bus and memory bus. Those cycles can be arranged in order within the CPU bus pipeline. A subset of cycles destined for a peripheral bus can be stalled within a snoop phase associated with the CPU bus. Snoop stall can continue until a memory cycle is encountered upon the CPU bus pipeline within a phase prior to the snoop phase. Once the memory cycle progresses to the snoop phase, snoop stall can be discontinued and the previous, peripheral cycles can then be deferred and/or retried, allowing the memory cycle to be quickly dispatched through all phases of the CPU bus and onto the memory bus.