 |
|||||||
|
|||||||
Peer-to-Patent - Archived ApplicationsUpdated: May 11, 2008 These descriptions have been prepared by the Peer-to-Patent project team and do not reflect the views of the inventors or their assignees. Application No: 11/425398 The use of pipelined resources in an SMP computer creates various timing/floor planning issues. To solve these issues, pipelines are often replicated and partitioned by using address/data steering techniques and parallel pipelines. Normally, these pipelines each use dedicated resources and the arbitration can occur independently within a pipeline. However, there are instances when multiple pipelines need to share common resources. Common resource sharing is currently accomplished in two ways. First, using an added shared queue to the common resources and having operations which exit each of the existing pipelines to be queued for the common resource. This can cost considerable queuing resources when both pipelines each need the common resource for many consecutive cycles and the logic to handle the process can be complex. Second, is to build extra arbitration into the top of each pipeline to avoid the case where more than one pipeline can contend for the same common resource. This can hurt timing on all requests to the pipelines causing a negative impact to performance. This invention eliminates the problems with the methods above by utilizing a post-arbitration reject step in each pipeline. Whenever there is a resource conflict within a pipeline after arbitration, a reject is issued to re-queue the request that had a conflict. Once the resource is free, the request can vie for priority back into the pipeline. Application No: 11/472929 Increasingly, people are required to remember passwords in order to access computer services. For some security applications, for example, it is desirable to have encryption keys as large as 128 bits or more. For many people, such random passwords are hard to remember, and typically the user will write the passwords down somewhere or otherwise store the passwords, compromising the security of the password. Current solutions assign one or more English code words to each bit pattern of a fixed number of bits. While this is easier to memorize than a 128 bit key, the generated mnemonic is still a random word sequence where the words have no relationship to each other. The application describes a method for generating mnemonic random passcodes comprised of choosing a schema to guide the generation of the word sequence and then transforming the passcode into the word sequence using the schema. Application No: 11/451751 The physical memory of a computing system is usually divided into physical pages. In computing systems that employ a virtual memory management scheme, some of the physical pages are mapped into the system kernel's virtual memory address space. Furthermore, a computing system may comprise multiple system boards, each board having its own range of physical memory addresses. A large computing system may be logically divided into multiple separate domains with separate functions (i.e. one domain to run a web server, another to run a database, etc.), each of which might be allocated one or more system boards. It is sometimes desirable to change the allocation of system boards. Reallocating the system boards entails relocating the data stored on that system board's physical pages to another system board's physical pages, a time consuming and complicated process. Because of the likelihood that system boards in a large computing system will be reallocated from time to time, all kernel pages are confined to a limited subset of all of the system boards. However, this limited subset of system boards containing kernel pages are often subject to heavy contention and thus the overall performance of the entire computing system may be degraded. The application describes a technique for relocating the contents of kernel pages, whereby kernel pages may be distributed among all computing system boards Application No: 11/421724 Videoconferencing and other forms of virtual collaboration allow the real-time exchange of video, audio, and other data among systems in remote locations. Typically this occurs over a computer network in the form of streaming video and audio data. Many systems can establish media streams at the beginning of an event, but cannot transition smoothly to new configurations as various systems enter or leave an event. While numerous methods have been devised to connect systems with identical or substantially compatible capabilities, managing events involving systems with differing capabilities is substantially more difficult. Application No: 11/497886 Videoconferencing and other forms of virtual collaboration allow the real-time exchange of video, audio, and other data among systems in remote locations. Typically this occurs over a computer network in the form of streaming video and audio data. Many systems can establish media streams at the beginning of an event, but cannot transition smoothly to new configurations as various systems enter or leave an event. While numerous methods have been devised to connect systems with identical or substantially compatible capabilities, managing events involving systems with differing capabilities is substantially more difficult. As current virtual collaboration systems do not adequately support systems with varying levels of functionality or allow dynamic reconfiguration of participating systems without interruption of an event in progress, the application describes a system and method for managing virtual collaboration systems. This includes a communication network coupled to a plurality of nodes and a gateway subsystem, with the gateway subsystem in turn coupled to a management subsystem. The management system is configured to dynamically manage the topology of a virtual collaborative event based on the node signals. Application No: 11/443154 A computer network may comprise hundreds or thousands of interconnected devices such as routers, switches, and end-nodes that may include computers, servers, and printers. Because of the sheer number of end-nodes, it is an end-node that typically causes network problems due to faulty hardware, virus infestation, and/or too much traffic due to the uploading or downloading of files over the network. The application describes a method and apparatus for determining the switch port to which an end-node device is connected. This would allow a user to receive an address for the end-node being used and discover a switching device that forwards traffic only to the selected end-node. Application No: 11/321022 Computer processing units (CPUs) are fed instructions in order to perform an operation. Typically, the instruction tells the CPU to read data and then manipulate the data in order to reach an outcome. However, the instruction rarely contains the actual data that the CPU needs. Rather, it contains an address, or "pointer" to the relevant data. Modern CPUs implement a technique called instruction pipelining that is similar to an assembly line. In simplified terms, one CPU sub-unit might decode the pointer, with the next sub-unit fetching the data, and finally one sub-unit might perform the operation. Instead of merely pipelining the instructions, vector processors also pipeline the necessary data. Thus, instead of having to decode every address that holds relevant data, the vector processor knows from the instruction that there is a series of addresses holding relevant data and it can then accurately predict what the next address will be. While the pipelining of instructions and data can provide significant time-savings, vector processors are designed to have a specific data width and if the execution data path does not match a maximum vector length, micro operations contained within the instructions may be ported to code using a shorter data length, causing the code to run slower. The patent application discloses a method for a Vector length tracking mechanism whereby a vector length tracker receives a value and then establishes a vector length for subsequent micro-operations to be executed corresponding to the value. Application No: 11/437223 A computer network may connect many devices such as desktop computers, printers, web servers, routers, databases, and laptops. In a large networked environment these devices are routinely being connected and disconnected. In such an environment it is difficult to accurately know what software may be loaded on each device and what devices are connected to the network at any given moment. A large networked environment may create a risk of having networked devices connected without the knowledge or permission of network managers. Unauthorized networked devices may contain viruses, lack proper virus protection, or may be used for unauthorized capture of network traffic. A need has arisen for network managers to be updated about unauthorized networked devices within the shortest amount of time and what software may be loaded on each network device. This application describes a system and method to dynamically scan a network to discover unauthorized devices. It uses a first network scanner and a second network scanner wherein a new network address is identified by the second network scanner but not the first network scanner. Application No: 11/419690 Manufacturing applications are used for controlling and monitoring processes. The typical user interface environments for these applications utilize a graphical tree control to organize and present data to a user. The user selects an item of interest from the graphical tree similar to a folder system in which all items have parent and child objects forming branches that can be expanded and collapsed to view or hide features for selection. Tree control software systems then present menus within the selected item for selecting an operation. There are many problems with a tree control system user interface for these applications. Accordingly, a need exists for a new user interface incorporating a paradigm which enhances the overall user experience. This application describes a user interface for manufacturing applications that employs a paradigm having a single progression for displaying manufacturing system information regardless of the functional point of view of the user. The user interface simplifies the navigation to particular items of interest and provides components that allow immediate and visible access to the operations that can be performed on the selected items. Application No: 11/437223 A computer network may connect many devices such as desktop computers, printers, web servers, routers, databases, and laptops. In a large networked environment these devices are routinely being connected and disconnected. In such an environment it is difficult to accurately know what software may be loaded on each device and what devices are connected to the network at any given moment. A large networked environment may create a risk of having networked devices connected without the knowledge or permission of network managers. Unauthorized networked devices may contain viruses, lack proper virus protection, or may be used for unauthorized capture of network traffic. A need has arisen for network managers to be updated about unauthorized networked devices within the shortest amount of time and what software may be loaded on each network device. This application describes a system and method to dynamically scan a network to discover unauthorized devices. It uses a first network scanner and a second network scanner wherein a new network address is identified by the second network scanner but not the first network scanner. Application No: 11/419690 Manufacturing applications are used for controlling and monitoring processes. The typical user interface environments for these applications utilize a graphical tree control to organize and present data to a user. The user selects an item of interest from the graphical tree similar to a folder system in which all items have parent and child objects forming branches that can be expanded and collapsed to view or hide features for selection. Tree control software systems then present menus within the selected item for selecting an operation. There are many problems with a tree control system user interface for these applications. Accordingly, a need exists for a new user interface incorporating a paradigm which enhances the overall user experience. This application describes a user interface for manufacturing applications that employs a paradigm having a single progression for displaying manufacturing system information regardless of the functional point of view of the user. The user interface simplifies the navigation to particular items of interest and provides components that allow immediate and visible access to the operations that can be performed on the selected items. Application No: 11/382364 Enterprise computing refers to the networks and computing systems used in large, often widely dispersed, corporations. Enterprise computing monitoring relates to the monitoring of the different, granular components of an enterprise computing environment. They are typically organized in a hierarchical fashion, with sensors distributed throughout the network. The sensors relay monitored events that can be interpreted, translated, and provided to interacting administrators in order to facilitate the management of the enterprise computing. Due to the complexity and magnitude of many enterprise computing systems, plainly reporting every event arising in hierarchy can result in an event flood that can quickly overwhelm the monitoring system. Because capturing every event stemming from a root cause is not as helpful as correcting the root cause, the focus is on resolving the root cause as that invariably leads to resolution of all other resulting events. The application describes a method, system, and computer program product for detecting event correlation rules in an enterprise computing monitoring system. This includes determining temporal proximity of events and frequency of occurrence, as well as reporting causal relationships of events. Application No: 11/416647 Computer systems and electronic devices, such as microprocessors, may contain timer circuits to maintain a running numerical count for use by other logic or devices. For example, the timer can be used to count time between events or maintain a real-time clock. On occasion, a timer may need to be updated with a new value by a user or computer program. A known method of updating a timer is by first reading the current timer value, storing it, loading a new timer value, and then programming the new timer value into the timer. The problem is that these operations take time to perform. Therefore, some time may pass between the first operation and when the new timer value is finally programmed into the timer. This creates a difference between the “real time” and the time reflected by the timer. Each timer update increases the difference between real time and the timer value. This can cause problems for programs or processes running on the computer. This application describes a technique to modify a timer without the timer advancing by a significant amount. In one embodiment, the time corresponding to the amount the timer increments (decrements) may be added (subtracted) to (from) the update timer value to compensate for elapsed time during the update process. In other embodiments, the number of operations used to update the timer is reduced in order to reduce the elapsed time during the update process. Application No: 11/380939 Wind turbines for wind power generation are typically grouped in large numbers referred to as wind-farms. Large wind-farms can comprise wind turbines numbering in the hundreds. Within each wind turbine are a number of possible devices including controllers and sensors. Typically, wind-farms have a centralized control unit connected to a local area network of wind turbines to allow for monitor and control of the entire wind-farm. Each device connected to the wind-farm network requires an IP address in order to communicate on the network using internet protocol. In addition, the central control unit needs to be able to uniquely identify each device in this network. A unique identification is lost using dynamic address assignment. For this reason, network settings in a wind-farm network are done manually. This is extremely time consuming, labor intensive, and prone to error. This application describes a method of automatic configuration of a local wind-farm network. The method operates by automatically assigning an IP address to a device connected to the network, automatically receiving an identifier tag from the device, and relating that automatically assigned IP address to a specific device using the identifier tag. Application No: 11/739021 The claimed invention is directed toward improving the security of protected assets, particularly assets protected with a password system, including personal identification numbers. In spite of the many efforts made to increase the security of password-accessible systems, there remains a need to provide more flexible convenient systems in which users can configure security rules for access to secured assets. The application describes a security system for protecting assets, including a password-based security system that can provide different levels of access responsive to entry of a primary or secondary password. In some versions, user-configurable security rules can provide customized responses to entry of primary or secondary passwords including feigned or limited access, security alerts, etc. Passwords comprising overt and covert components can be used to provide enhanced security and improved user control over system response. Improved security systems involving transactions between multiple parties are also considered, with options for user-customized security rules including primary and secondary passwords, and reverse challenge and response methods. Systems for Limited Use Credentials are also disclosed to reduce the risk of identity theft. Application No: 11/405994 In computer applications that access data in a streaming fashion it is often desirable to estimate the sortedness of the data steam. Sortedness being the number of operations required to place the elements in the data steam in a sorted order. An estimation of sortedness may be useful in data steams corresponding to network routing or when comparing very long rankings (i.e. ordered lists of distinct items). However, it is difficult to estimate sortedness in a feasible manner without having to scan the data steam more than once and/or without having to actually sort the entire sequence. Therefore, there is a need for a method that requires only one pass over a sequence of data elements and adequately samples data elements from the sequence to allow for an estimation of sortedness. This application describes a method of scanning in a data stream once in order to obtain uniform data samples from selected intervals and using the obtained data samples to estimate the degree of sortedness of that data steam. Specifically, multiple samples are randomly selected for storage in one or more data buckets. Those samples are then randomly obtained from the bucket(s). Each sample is associated with a specified interval immediately prior to a selected point in time. The balance of probabilities between selection of elements stored in the bucket and the selection of elements included in the samples allows that elements scanned during the specified interval are included in the sample with equal probability. These multiple data samples are then used to estimate the degree of sortedness of the data stream. Application No: 11/393198 The ability to drag-and-drop objects in a user interface has been around for some time. It does however have shortcomings. When dragging an object to a drop target in a remote part of the interface, a user can often expect to make multiple gestures with the mouse or touchpad to get to the desired location. Similarly, if the drop target is in a part of a list that is offscreen, the user must wait and hold the object while the list scrolls to the target. The claimed invention provides mechanisms for improving the usability of drag-and-drop functionality. The application provides methods and apparatus for manipulating objects in a user interface. After a user selects an object to move, at least one additional interface object, representing a drop target with which the selected object may be associated, is presented in the user interface in proximity to the selected object. Application No: 11/395232 Crossbar arrays have been used in a number of applications, primarily in information storage and retrieval. While there has been some development in using crossbar topology to replicate the functions of the basic logical units (AND, OR, NAND, etc.) as used in conventional arithmetic logic units, the claimed invention uses a different approach to provide a new level of computational power. The application provides for a crossbar arithmetic processor combining a crossbar array with a programming unit, an input unit, and a post procession unit. The programming unit sets the conductive states of the crossbars, which are preferably controlled through rectification layers between layers of the crossbar array. The input unit provides a bit pattern of input numerical values to the columns while the post-processing unit converts analog signals output from the rows of the crossbar array into digital output bit patterns that are further processed into an output numerical value. The end product is an arithmetic sum or product based on the programmed and input numerical values. Application No: 20070234286 Thunking describes any variety of process by which a first process executing a first set of machine executable instructions compiled with specific parameter (platform type, word size, bit-value, etc.) is able to successfully make a function call to another set of machine executable instructions compiled for another set of parameters. In general, thunking involves a translation of function calls, function call parameters, and/or return parameters between at least two processes compiled for differing specifications. When a process makes a function call to another process being executed by a virtual machine, the virtual machine will copy calling parameters from the initial process. The calling parameters in many cases will be actual data needed by the virtual machine in order to complete the request. In other cases, the calling parameters may be pointers to data. An increasingly common trend (as in the Intel® Extended Memory 64 Technology) has the processor send either actual data or data pointers based on the size of the data structure being sent. Currently, virtual machines have no way of knowing whether the passed data structure is actual data or a pointer to data and thus the virtual machine cannot successfully implement the necessary thunking. The application describes methods and apparatus to implement annotation based thunking. The process sends as call paramaters, based upon a predetermined threshold, either data or a data pointer and adds to the data structure an annotation record to allow the process receiving the function call to complete the request. Application No: 11/389579 The application provides for a system that selectively executes different versions of executable code for the same source code. The system receives an executable code module containing two or more versions of executable code, each optimized for a different condition. Upon receiving the executable code module, the system administers and evaluates a test condition. After evaluating the test condition, the system executes the version of executable code that is optimized for the test condition. Application No: 11/388524 Companies looking to reduce the high cost of data storage often aggregate data onto shared virtualized storage systems thereby reducing infrastructure and management overhead of storage systems. However, the aggregation of data onto shared virtualized storage systems can lead to unexpected interference between applications using the same storage device. Therefore, a virtualized storage system must provide assurance that the behavior of one application does not interfere with the performance of other applications. Conventional storage systems manage resources according to specification of reserves and limits. Virtualized storage systems manage resources in pools and sessions. A pool represents the virtual entity composed of reserve and limit resources used by an application. Within each pool, each application may subdivide into sessions. The problem of managing I/O performance resources can be divided into the problems of (1) how to specify allocations for pools and sessions and (2) how to deliver on those allocations. There are existing approaches to resource management which enforce fair sharing between applications and enforce reserves and limits; however, there is a need for a performance resource management system that further prioritizes applications based on deadlines. This application relates to an invention that satisfies this need, and presents a system, a service, a computer program product, and an associated method for managing storage system performance as a resource.
Application Number: 11/389857 Clinical decision support systems provide assistance to physicians and other healthcare providers in making decisions regarding diagnosis and/or treatment of a patient. In order to provide valuable assistance, a clinical decision support system must utilize relevant medical evidence about a patient. Medical evidence, such as, patient medical histories, imaging data, test results, allergies, and other reports, can be stored on various electronic medical systems which may be located in more than one geographic location. With current clinical decision support systems, the healthcare practitioner must manually seek out medical evidence to be used with the system. This application relates to a system, method, and set of instructions on a computer-readable medium used to search a plurality of medical evidence systems for medical evidence given a search parameter.
Application Number: 11/304021 The application discloses an invention that provides a displayable progress indicator for a program execution by a corresponding computing device. The progress indicator is updated to accurately reflect the progress of the program execution by the computing device. This invention discloses a method for updating and inserting a progress indicator of a multi-step program execution by a computing device. This method involves a computation for each step of the multi-step program based on a completion time baseline for each step of the multi-step program. The method would apply to any multi-step application or operating system program that is executable in steps by any type of computing device. The progress indicator can have any form including the form of a progress bar.
Application Number: 11/276483 Computers browsing internet sites are often exploited by the sites they visit. Internet sites can either perform actions not requested by the visiting computer or even trick the user into performing an un-intended action. A typical example of such an exploit is the installation of malware. This application relates to a system/method used to actively explore sites on a network and detect if a site is capable of accomplishing exploits on a machine. This “honey monkey system” detects malicious browser-based exploits by actively searching uniform resource locators (URL’s) on a network and uses a browser-based vulnerability detector to detect exploits.
Application Number: 11/353478 The application relates to a system for implementing synchronized objects in software transactional memory comprising one or more processors. It also provides for a memory comprising program instructions executable by the processor in order to implement a transactional-memory manager that is configured to coordinate memory access requests directed at the memory from a plurality of transactions.
Application Number: 11/276178 The application describes a method for operating a gateway for wireless mobile clients in a messaging system, a core messaging system, and at least one gateway process or gateway acting as a proxy on behalf of the clients. Application Number: 11/333019 The application describes a system of security management for software applications associated with multiple user registries that includes an integrated console configured to host one or more applications or resource objects in corresponding realms. Application Number: 11/342055 Many data management systems are implemented as a single centralized application, which users access via the world wide web. Advantageously, this implementation minimizes a burdensome need to provide individualized application service, maintenance, and upgrading at local levels. If, however, an application server, web server, or the like associated with a centralized application needs to be restarted for some reason, then it is likely that at least some of its users may endure information and data loss or session termination, which are undesirable, inconvenient, and commonly require users to re-log into the application in order to continue use. This application describes a system and method whereby information is still stored on an external database but is retrieved into a memory associated with a server. The user is then able to access the memory through an interface. Users are able to maintain user sessions and prevent information and data loss or session termination resulting from service interruptions to a data management system. Application Number: 11/312201 Many computer systems involved in mission critical activities (such as banking, telecom, and stock markets) depend upon a high level of computer reliability. In order to achieve reliable, fault- tolerant computing, many computer systems utilize multiple processors. Processors operate at core voltages set by the manufacturer. When multiple processors are used, the resulting nondeterministic behavior can lead to system failures. This application describes a method, apparatus, and system for tuning core voltages of processors in order to avoid failures. Application Number: 10/548823 Complex systems use a multitude of processes, which are linked to one another, making it very difficult to predict the influence of changes of individual links and parameters on the behavior of the complete system. Since the number of influencing variables increases exponentially with the size of a manufacturing process, company, or project, one may not be able to distinguish relevant from non-relevant variables. Currently, models are directed to a certain, current system condition, which permits an evaluation of current or past characteristic values for assessing the system performance and there exists no possibility of comparing different systems with one another. This application describes a method, a data-processing system and a computer program for the comparison of models of a complex system. Application Number: 11/329801 This application describes a system whereby local workstations provide screen update data, which is received by a remote console that is enabled to display the data from at least two of the local workstations simultaneously as well as interface with the local workstations. The invention relates to the simultaneous observation and/or operation of a plurality of workstations by remote console. The invention has application to a number of fields, such as providing support for medical imaging systems. Rather than retaining a technician that is able to travel to each medical imaging system that needs support, a remote service technician may access the imaging system to perform diagnostic routines, configure imaging settings, and train local personnel to use the system. Previous remote observation systems allow a single connection between the local operator and the remote console resulting in a 1:1 ratio of training personnel to trainees. Hence, when a new tool or interface is released a trainer may have to repeat the same training for many different operators, which is time-consuming and expensive. Application Number: 11/329773 This application describes a cipher method for verifying a decryption of an encrypted user data key. It relates to methods for checking the validity of a decrypted user data key upon restore of encrypted user data in a manner that links the user data key and a key used to decrypt an encryption of the user data key. The invention is intended to meet the need for a secure and reliable method for verifying a correct decryption of the encrypted user data key in view of asymmetric key-based cipher methods. Application No: 11/329656 This application relates to the delivery of a device driver to an operating system without user intervention. While most computer platform operating systems are configured with device drivers for all devices installed in the computer platform when it was purchased or delivered, if a newer driver becomes available after the platform is delivered, the owner may need to download or install the driver from diskette, across a network, etc. This makes it difficult to execute an old operating system on a new computer platform because the operating system will generally not be configured with drivers for controlling all components of the new platform. Because there is no readily available method or scheme for automatically updating an old operating system with the drivers needed for a new computer platform (i.e. the driver was not available when the operating system was originally released), this application describes a method and apparatus whereby a device is identified for which a driver is needed during the boot phase of an operating system and the driver is then requested from a service processor of the computer platform, retrieved, and delivered to the operating system. Application Number: 11/296194 This application describes systems and methods for an off-line economy for digital media whereby media devices of buyer and seller perform secure off-line sales transactions that are recorded on the media device so that a percentage of the sale price can be applied to a copyright owner as well as the seller as an incentive. Thus, even resale of pirated media content benefits the copyright holder. This attempts to solve problems associated with both copyright holders who have been driven to seek their revenue on-line, which limits customers to purchasing media clips only when they are connected to the Internet, and the widespread phenomenon of file sharing, where content distribution is economically isolated from copyright holders. Application Number: 11/291378 This application describes methods and apparatus to identify memory communication by creating a stack tracker structure to store information that may be utilized to generate a distance value corresponding to a relative distance between a load instruction and a previous store instruction. The invention attempts to solve problems associated with memory renaming (MRn) which allows transformation of memory communication into register-register communication. MRn predicts which data will be renamed and instructions that source data that is predicted to be renamed can then source the data directly from the original producer instead of waiting for the data store to load memory communication. A correct prediction avoids load latency associated with accessing memory outside of the processor. When a prediction is incorrect, the processor state must be returned to a point prior to the misprediction and performance suffers. By generating a distance value between load instructions and store instructions, the invention attempts to minimize the number of incorrect predictions. Application Number: 11/304021 IBM's application describes an invention for optimizing the migration of a database by collecting details (parameters) about the source system and the target system in order to generate configuration information that can be used to fine tune the database's operation on the target system. This invention attempts to solve the problem presented by inconsistencies between the source and target systems that interfere with the migration of a database. Application Number: 11/301,399 This invention relates generally to a system for grouping and managing multiple process related variables for an industrial system. For example, when control room operators operating a power generation facility view processes associated with a system using a display type console, a large number of individual parameters are observed. These parameters are dynamic (or variable) and dependent upon aspects of the facility operations. Collectively, a certain set of variables is indicative of the status of the process. In operations parlance, each member of the set of variables is commonly referred to as a "tag." Evaluation of process status by viewing a set of tags can be confusing. This confusion can be exacerbated by the fact that a tag for one process can also relate to another process. This application claims an apparatus and a method for receiving operational data and displaying summaries of processes for an industrial system includes a block object to define blocks for defining aspects of variables related to the operational data as well as a display client for using the blocks to provide useful and concise summaries of the processes. Application Number: 11/286,554 This application relates to electronic computing, and more particularly to a computing system that implements user selectable management alert format. When a system manager remotely accesses or controls a computer system with devices attached to it, the devices may be able to alert the system manager of problems in a number of formats, with each format requiring a separate program to read the alert. For example, network interface cards (NICs) or other computing components may be operable in accord with the alert standard format (ASF) protocol or with the active management technology (AMT) protocol. In many instances, the devices may use common hardware, and the device is configured to operate in accordance with a specific management alert protocol though software or firmware installed onto the device. This application describes a system for improving efficiency of remote access by displaying to the user a list of formats available (as determined from the devices attached to the computer system being remotely accessed) and allowing the user to select a single format for all of the devices to use. Application Number: 11/290,893 Large number of accesses to datastores that retain information related to various business functions (i.e., business or medical records) can lead to degradation in the performance of an informational system. This application describes methods for improving datastore management information systems by enhancing the ability to process, store, and retrieve data quickly and more efficiently. The embodiments of this invention provide several methods and systems for managing an operational datastore and a staging area datastore. Application Number: 11/286,585 Prefetching in microprocessors can help to load instructions or data used by a processor ahead of when the instructions or data are actually needed, thereby improving performance. Performance suffers, however, if the prefetched data or instructions are later found not to be needed by the processor. At least one technique described in this patent application, enables numerous types of prefetch access patterns used by a processor to prefetch data and/or instructions, so as to help increase the number of prefetched data and/or instructions that are actually used by the processor during the course of program execution, while reducing the number of prefetched data and/or instructions that are not. Application Number: 11/164,393 The invention relates to techniques for increasing performance efficiencies of memory management. Existing physical memory allocation schemes result in fragmentation and cause inefficient re-use of physical memory. This in turn makes the computer system run slower. A solution is proposed which would enhance computer performance by making enhancements to the memory allocation routines. The method includes enabling a kernel to free memory in a physical memory space corresponding to memory allocations released by processes in a virtual memory space. After freeing the memory, the system determines whether freed physical memory in the physical memory space spans one or more fixed size memory units (e.g., page frames). The method further includes designating a status of the one or more page frames as available for reuse; the freed page frames marked as available for reuse being available for backing new processes without requiring the kernel to delete data included in the freed memory released by the process. According to the description of the invention, this is possible because “rather than recycling a page when no longer needed by a process, the kernel records the page as available for reuse…thus cutting overhead associated with recycling the available page.” |
|||||||
|
2.gif)
