199 results about "Multiprocessing" patented technology

A single chip protocol converter integrated circuit (IC) capable of receiving packets generating according to a first protocol type and processing said packets to implement protocol conversion and generating converted packets of a second protocol type for output thereof, the process of protocol conversion being performed entirely within the single integrated circuit chip. The single chip protocol converter can be further implemented as a macro core in a system-on-chip (SoC) implementation, wherein the process of protocol conversion is contained within a SoC protocol conversion macro core without requiring the processing resources of a host system. Packet conversion may additionally entail converting packets generated according to a first protocol version level and processing the said packets to implement protocol conversion for generating converted packets according to a second protocol version level, but within the same protocol family type. The single chip protocol converter integrated circuit and SoC protocol conversion macro implementation include multiprocessing capability including processor devices that are configurable to adapt and modify the operating functionality of the chip.

A mechanism for exception and interrupt handling in multithreaded multiprocessors is provided. The mechanism allows the handling of exceptions and interruptions in a multithreaded multiprocessor computer, while hiding the multiprocessor nature of the computer from the operating system. Generally, when an operating system is cognizant of the multiprocessor nature of a computer, additional overhead may be required when handling exceptions and interruptions. Due to the overhead involved in saving and restoring processing states, the performance of a processor may be significantly impacted. Additional circuitry is provided which allows the multiprocessor nature of the computer to be hidden from the operating system, while minimizing the overhead necessary for proper handling.

A point-to-point connected multiprocessing node uses a snooping-based cache-coherence filter to selectively direct relays of data request broadcasts. The filter includes shadow cache lines that are maintained to hold copies of the local cache lines of integrated circuits connected to the filter. The shadow cache lines are provided with additional entries so that if newly referenced data is added to a particular local cache line by “silently” removing an entry in the local cache line, the newly referenced data may be added to the shadow cache line without forcing the “blind” removal of an entry in the shadow cache line.

Cache sharing for a chip multiprocessor. In one embodiment, a disclosed apparatus includes multiple processor cores, each having an associated cache. A control mechanism is provided to allow sharing between caches that are associated with individual processor cores.

An infrastructure and a set of steps are disclosed for evaluating performance of computer systems. The infrastructure and method provide a flexible platform for carrying out analysis of various computer systems under various workload conditions. The flexible platform is achieved by allowing/supporting independent designation/incorporation of a workload specification and a system upon which the workload is executed. The analytical framework disclosed and claimed herein facilitates flexible/dynamic integration of various hardware models and workload specifications into a system performance analysis, and potentially streamlines development of customized computer software/system specific analyses.

The disclosed performance technology infrastructure includes a workload specification interface facilitating designation of a particular computing instruction workload. The workload comprises a list of resource usage requests. The performance technology infrastructure also includes a hardware model interface facilitating designation of a particular computing environment (e.g., hardware configuration and/or network/multiprocessing load). A disclosed hardware model comprises a specification of delays associated with particular resource uses. A disclosed hardware specification further specifies a hardware configuration describing actual resource elements (e.g., hardware devices) and their interconnections in the system of interest. The performance technology infrastructure further comprises an evaluation engine for performing a system performance analysis in accordance with a specified workload and hardware model incorporated via the workload specification and hardware model interfaces.

A method (and system) of transparent dynamic optimization in a multiprocessing environment, includes monitoring execution of an application on a first processor with an execution monitor running on another processor of the system, and transparently optimizing one or more segments of the original application with a runtime optimizer executing on the another processor of the system.

The present invention is related to the processing of data, and more particularly to a method of and system for processing large volumes of Earth observation imagery data. A system for processing a large volume of Earth observation imaging data is described, comprising a computer including a visual display and a user interface, a plurality of servers, an image database storing said Earth observation imaging data as a plurality of separate image data files, and a network for interconnecting the computer, plurality of servers and image database. The plurality of servers is operable to process the separate data files in a distributed manner, at least one of the plurality of servers is operable to process the separate data files in a multiprocessing environment and at least one of the plurality of servers is operable to collate the processed separate data files into a single imaging result.

A query parser that converts query definition into a query graph and decides the execution order of operators is installed, a set of consecutive operators in the execution order is called a stage, the total of calculation costs of operators configuring each stage is called a calculation cost of the stage, the query graph is divided into multiple stages such that the calculation cost of each stage becomes a value less than a value dividing the total cost of all operators by the number of calculation cores, and each calculation core extracts tuples one by one from an input stream, and, when taking charge of and executing processing of the tuples from the entrance to exit of the query graph, before the execution of each stage, confirms whether processing of the stage is completed for a tuple previous to a charge tuple.

Methods and apparatus provide for: decrypting a first of a plurality of operating systems (OSs) within a first processor of a multiprocessing system using a private key thereof, the plurality of OSs having been encrypted by a trusted third party, other than a manufacturer of the multiprocessing system, using respective public keys, each paired with the private key; executing an authentication program using the first processor to verify that the first OS is valid; and executing the first OS on the first processor.

In a multiprocessing system, a configuration manager maintains various reference parameters that are selectively copied by subordinate managed units to form local operating parameters, which subsequently govern operation of these managed units. A comparative technique is employed to track reference parameter updates, and synchronize each local operating parameter counterpart accordingly. At the configuration manager, reference parameters include reference profiles and reference characteristics. Each reference profile specifies one or more of the reference characteristics. At each managed unit, the operating parameters include subcribed-to profiles and operating characteristics; both are initially copied from the configuration manager's reference profiles/characteristics. Each local operating profile specifies one or more of the operating characteristics. Each managed unit operates according to its locally maintained operating characteristics. When certain update criteria are satisfied, a managed unit and the configuration manager cooperatively synchronize the manage unit's local operating profiles and characteristics with the configuration manager's reference profiles and characteristics. This involves comparing the reference and operating profiles to identify new, updated, or deleted operating characteristics. Also, the local operating profiles and operating characteristics may be cross-referenced to identify any "orphan" characteristics for deletion.

In a multiprocessing system, hierarchically superior configuration managers maintain profiles of operating characteristics to which subordinate managed units selectively subscribe. If the profiles or operating characteristics change, the configuration managers propagate the changes to all managed units. Each configuration manager stores a record of operating characteristics and multiple server profiles, each profile specifying one or more operating characteristics. A subscription list identifies one or more managed units, each associated with one or more server profiles. Each managed unit acts according to its current operating characteristics, stored locally at the managed unit. If the managed unit receives a profile subscription request from a system administrator, the managed unit sends a subscription message to the configuration manager to subscribe to that input profile. Receiving the subscription, the configuration manager enters the subscribing managed unit and the associated profile into the subscription list, and returns the profiled operating characteristics to the subscribing managed unit. The subscribing managed unit stores these operating characteristics in its record of current operating characteristics. If there is a change to the operating characteristics (or to the profiles), the configuration manager transmits the changed matter to all managed units with affected subscriptions. Upon receipt of this data, each subscribing managed units stores the changed operating characteristics in its record of current operating characteristics.

In a multiprocessor system it is generally assumed that peak or near peak performance will be achieved by splitting computation across all the nodes of the system. There exists a broad spectrum of techniques for performing this splitting or parallelization, ranging from careful handcrafting by an expert programmer at the one end, to automatic parallelization by a sophisticated compiler at the other. This latter approach is becoming more prevalent as the automatic parallelization techniques mature. In a multiprocessor system comprising multiple heterogeneous processing elements these techniques are not readily applicable, and the programming complexity again becomes a very significant factor. The present invention provides for a method for computer program code parallelization and partitioning for such a heterogeneous multi-processor system. A Single Source file, targeting a generic multiprocessing environment is received. Parallelization analysis techniques are applied to the received single source file. Parallelizable regions of the single source file are identified based on applied parallelization analysis techniques. The data reference patterns, code characteristics and memory transfer requirements are analyzed to generate an optimum partition of the program. The partitioned regions are compiled to the appropriate instruction set architecture and a single bound executable is produced.

To assure the multiprocessing performance of CPU on a microprocessor, the invention provides a method of memory mapping for multiple concurrent processes, thus minimizing cache thrashing. An OS maintains a management (mapping) table for controlling the cache occupancy status. When a process is activated, the OS receives from the process the positional information for a specific part (principal part) to be executed most frequently in the process and coordinates addressing of a storage area where the process is loaded by referring to the management table, ensuring that the cache address assigned for the principal part of the process differs from that for any other existing process. Taking cache memory capacity, configuration scheme, and process execution priority into account when executing the above coordination, a computer system is designed such that a highest priority process can have a first priority in using the cache.

A method and apparatus for managing the execution on guest processors of a broadcast instruction requiring a corresponding operation on other processors of a guest machine. Each of a plurality of processors on an information handling system is operable either as a host processor under the control of a host program executing on a host machine or as a guest processor under the control of a guest program executing on a guest machine. The guest machine is defined by the host program executing on the host machine and contains a plurality of such guest processors forming a guest multiprocessing configuration. A lock is defined for the guest machine containing an indication of whether it is being held by a host lock holder from the host program and a count of the number of processors holding the lock as guest lock holders. Upon decoding a broadcast instruction executing on a processor operating as a guest processor, the lock is tested to determine whether it is being held by a host lock holder. If the lock is being held by a host lock holder, an instruction interception is recognized and execution of the instruction is terminated. If the lock is not being held by a host lock holder, the lock is updated to indicate that it is being held by the guest processor as a shared lock holder, the instruction is executed, and then the lock is updated a second time to indicate that it is no longer being held by the guest processor as a shared lock holder.

A system, method and computer program product for processing overflow transactions in a transactional memory system. The transactional memory system is provided in a multiprocessing system having one or more processor devices and a shared memory storage system, and implements a best effort hardware transactional memory system. The method includes acquiring, by a requesting processor, lockbits associated with a memory structure of the shared memory storage system to be reserved for an overflowing transaction. The lockbits determine the granularity at which memory reservations for an overflow transaction are recorded. The method includes implementation of control mechanism for controlling concurrency between overflowing and non-overflowing transactions requested by processor devices in the multiprocessing system, the method enabling only one overflowing transaction to execute at a time in the multiprocessing system.

A computer implemented method, apparatus, and computer usable medium for gathering performance related data in a multiprocessing environment. Instrumentation code is executed on a processor that minimizes the distortion to the processor resources used to execute the program to be profiled. Data is written by the instrumentation code to a shared memory in response to an event occurring during execution of the program. The data is generated during execution of the program on the processor and the instrumentation code uses shared memory to convey the data to a profiling application running on a set of profiling processors. The data is collected by the set of profiling processors in the shared memory written by the instrumentation code.

Embodiments of the invention provide methods and systems for reducing the consumption of inter-node bandwidth by communications maintaining coherence between accelerators and CPUs. The CPUs and the accelerators may be clustered on separate nodes in a multiprocessing environment. Each node that contains a shared memory device may maintain a directory to track blocks of shared memory that may have been cached at other nodes. Therefore, commands and addresses may be transmitted to processors and accelerators at other nodes only if a memory location has been cached outside of a node. Additionally, because accelerators generally do not access the same data as CPUs, only initial read, write, and synchronization operations may be transmitted to other nodes. Intermediate accesses to data may be performed non-coherently. As a result, the inter-chip bandwidth consumed for maintaining coherence may be reduced.

A system and method are provided to dedicate one or more processors in a multiprocessing system to performing encryption functions. When the system initializes, one of the synergistic processing unit (SPU) processors is configured to run in a secure mode wherein the local memory included with the dedicated SPU is not shared with the other processors. One or more encryption keys are stored in the local memory during initialization. During initialization, the SPUs receive nonvolatile data, such as the encryption keys, from nonvolatile register space. This information is made available to the SPU during initialization before the SPUs local storage might be mapped to a common memory map. In one embodiment, the mapping is performed by another processing unit (PU) that maps the shared SPUs' local storage to a common memory map.

Lock-free circular queues relying only on atomic aligned read/write accesses in multiprocessing systems are disclosed. In one embodiment, when comparison between a queue tail index and each queue head index indicates that there is sufficient room available in a circular queue for at least one more queue entry, a single producer thread is permitted to perform an atomic aligned write operation to the circular queue and then to update the queue tail index. Otherwise an enqueue access for the single producer thread would be denied. When a comparison between the queue tail index and a particular queue head index indicates that the circular queue contains at least one valid queue entry, a corresponding consumer thread may be permitted to perform an atomic aligned read operation from the circular queue and then to update that particular queue head index. Otherwise a dequeue access for the corresponding consumer thread would be denied.