Abstract: Processing of an instruction fetch from an instruction cache is provided, which includes: determining whether the next instruction fetch is in a same cache line of the instruction cache as a last instruction fetch; and based, at least in part, on determining that the next instruction fetch is in the same cache line, suppressing for the next instruction fetch one or more instruction cache-related directory accesses, and forcing for the next instruction an address match signal for the same cache line. The suppressing may include generating a known-to-hit signal where the next fetch is in the same cache line, and the last fetch is not a branch instruction, and issuing an instruction cache hit where a cache line segment of the same cache line having the next instruction has a valid validity bit, the valid validity bit having been retrieved and maintained based on a most-recent, instruction cache-directory-accessed fetch.

Abstract: A method for operating a home automation system includes receiving a device modification request from a requesting user device of a requesting user to modify a configurable device property of a requested household device. The requesting user is one of one or more users associated with the household. The method further includes identifying the requested household device based on a home structure mapping the household devices to the corresponding device rooms based on a corresponding location of each household device in the residence. The method also includes determining whether the requesting user has an assigned access level granting access to the requested household device and allowing modification of the one or more configurable device properties of the requested household device. When the requesting user has been assigned an access level, the method includes permitting the modification of the at least one configurable device property of the requested household device.

Abstract: A method and system for the loading of an operating system on a computing device having a central processing unit and a dedicated operating system graphic processing unit, which provides Image read/write functionality to the central processing unit. The method comprises acquiring an image representation of the operating system, the image being composed of pixels having color values encoding operating system executable code to be executed by the central processing unit, providing the operating system image to the dedicated graphic processing unit and accessing the encoded operating system executable code using the coordinates of the image pixels.

Abstract: Embodiments here include systems and methods for running an application via a microvisor processor in communication with a memory and a storage is disclosed. For example, one method includes installing an application. The method also includes identifying a resource required by the application to execute, wherein the resource is part of an environment. The method also includes identifying a location of the resource in the storage. The method also includes retrieving the resource from the storage. The method also includes bundling the application and the resource in the memory. The method also includes executing the application using the resource.

Abstract: A technique for shielding real-time workloads from operating system (OS) jitter due to expedited read-copy update (RCU) grace periods. In accordance with the disclosed technique, a kernel parameter is set to indicate that expedited RCU grace periods are to be suppressed. The kernel parameter is checked to see if it is set. A normal non-expedited RCU grace period is invoked in response to attempted invocation of an expedited RCU grace period if the check reveals that the kernel parameter is set.

Abstract: Automated provisioning of hosts on a network with reasonable levels of security is described in this application. A certificate management service (CMS) on a host, one or more trusted agents, and a public key infrastructure are utilized in a secure framework to establish host identity. Once host identity is established, signed encryption certificates may be exchanged and secure communication may take place.

Abstract: Methods, systems, and media for presenting a user interface customized for a predicted user activity are provided. In some embodiments, the method comprises: selecting users of a content delivery service, causing user devices to prompt the associated users to provide subjective data related to the user's intent when requesting media content items, training a predictive model to identify a user's subjective intent in requesting a media content item based on objective data received from a user device associated with the user and the subjective data received from the user devices, wherein the predictive model is trained to identify whether to present the user with a first user interface associated with a first user intent or a second user interface associated with a second user intent, causing the first user interface or the second user interface to be presented.

Abstract: A computer-implemented method for recording a browsing session includes receiving data generated by at least one input device during the browsing session. The received generated data is relative to each individual action applied by the at least one input device in accordance with each specific visited webpage during the browsing session. The received generated data is converted with respect to each individual action into a textual format, such that the individual actions applied with respect to each specific visited webpage forms a unique recorded browsing session. The converted data is stored as a unique recorded browsing session in a database, such that the unique recorded browsing session can be recreated or reproduced, e.g., either in-live or on-demand all over again by a web browser or other player.

Abstract: The present disclosure discloses a method for customizing a desktop launcher of a mobile terminal, comprising: placing a resource file and a configuration file for customizing at least two kinds of desktop launchers in a resource folder and a configuration folder, respectively, wherein the files and the folders of the desktop launchers are different in name; adding a compilation command corresponding to the at least two kinds of desktop launchers, and copying the resource file and the configuration file corresponding to the at least two kinds of desktop launchers in a saving path to a second file directory; adding the saving path of the files corresponding to the at least two kinds of desktop launchers in the second file directory to a program information file; compiling the files in the folders corresponding to the desktop launchers in the second file directory, so as to generate an installation program APK file; and loading the installation program file corresponding to the desktop launcher selected by the

Abstract: An electronic device is provided. The electronic device includes a management module configured to determine if a package related to an application is equipped in the first electronic device, a communication module configured to receive determine if a package related to an application is equipped in the first electronic device, and a display configured to output an item for executing the application, wherein the management module controls the display to display the item based on a comparison of the package and the application information.

Abstract: A method for executing a plurality of child kernels invoked on a device side is provided. The child kernels are invoked in response to a parent kernel launched from a host side. The method includes the following steps: linking the child kernels to enqueue a plurality of threads of the child kernels; regrouping the threads of the child kernels to generate a plurality of thread blocks each having N threads, wherein N is a positive integer greater than one; merging the thread blocks to generate a consolidated kernel; and executing the consolidated kernel on the device side to execute a kernel function of the child kernels.

Abstract: A computer system receives an account creation request for a social media platform created and sent using a frontend component. An application programming interface (API) call sequence associated with the account creation request is received from the frontend component. The API call sequence can reflect API calls registered by the frontend component in connection with creation of the account creation request, and timings of the registered API calls. An API call sequence model is applied to the API call sequence. The API call sequence model can be generated by providing training API call sequences to a machine learning component. At least one action is taken in response to the application of the API call sequence model indicating that the API call sequence is anomalous. The action can be taken with regard to the account creation request, or with regard to an account created in response to the account creation request.

Abstract: An application is deployed on a first cloud from a set of clouds in a hybrid cloud environment. A simulator is deployed on each cloud in the set. A vertical relationship function is computed between a time-series of the application and a time-series of the simulator on the first cloud. A first actual value in the time-series of the application is forecasted for a future time. A horizontal relationship function is computed between a first simulator value in the time-series of the simulator on the first cloud and a second simulator value in the time-series of the simulator on a second cloud. A second actual value in the time-series of a hypothetical deployment of the application on the second cloud is forecasted for the future time without deploying the application on the second cloud. The application is placed on the second cloud when the second actual value satisfies a condition.

Abstract: A technique for scaling an application with a set of one or more virtual machines is described. The technique may be performed during runtime of the application and responsive to a determination that a scaling operation is required for the application. That determination can be based on at least one first performance measurement result for the application. A method implementation of the technique comprises calculating a scaling magnitude for the required scaling operation taking into account at least one second performance measurement result for the application. The scaling magnitude is indicative of a resource quantity to be added to or removed from the application. The method further comprises triggering generation of a scaling request. The scaling request is directed at a scaling of the application on the basis of the calculated scaling magnitude.

Abstract: A web server uses selection criteria for identifying an electronic consumer device to identify a codeset likely to be appropriate for communicating with the electronic consumer device and data indicative of at least one code of the identified codeset is provided to a remote control device whereupon the data indicative of the at least one code of the identified codeset is automatically activated by the remote control device.

Abstract: An on-demand disposable virtual work system that includes: a virtual machine monitor to host virtual machines, a virtual machine pool manager, a host operating system, a host program permissions list, and a request handler module. The virtual machine pool manager manages virtual machine resources. The host operating system interfaces with a user and virtual machines created with an image of a reference operating system. The host program permissions list may be a black list and/or a white list used to indicate allowable programs. The request handler module allows execution of the program if the program is allowable. If the program is not allowable, the host request handler module: denies program execution and urges a virtual machine specified by the virtual machine pool manager to execute the program. The virtual machine is terminated when the program closes.

Abstract: Examples may include intercepting packets outputted from a primary virtual machine (PVM) hosted by a first server and converting one or more fields of protocol headers for each intercepted packet such that output-packet-similarity may be increased between the PVM outputted packets and packets outputted by a secondary virtual machine (SVM) hosted by a second server.

Abstract: System and methods for creating and configuring virtual machines (VMs) are provided. The method may include receiving, a first VM configuration and a second VM configuration. The second VM configuration is related to the first VM configuration. The method may include creating a first plurality of VMs in a first VM group based on the first VM configuration and creating a second plurality of VMs in the first VM group based on the second VM configuration. The method may further include applying the affinity rule in the first VM configuration to the first plurality of VMs and the second plurality of VMs. The method may further include configuring the first plurality of VMs based on the set of VM services in the first VM configuration and the second plurality of VMs based on the set of VM services in the second VM configuration.

Abstract: A computer-implemented method includes: receiving, by a computing device, a template for creating a virtual machine (VM) instance; separating, by the computing device, the template into a repeated portion and a unique portion; determining, by the computing device, whether the repeated portion is stored in a cache; creating, by the computing device and based on determining that the repeated portion is stored in the cache, the VM instance using the repeated portion stored in the cache; completing, by the computing device, the unique portion of the VM instance to create a completed VM instance; and deploying, by the computing device, the completed VM instance.

Abstract: A migration system includes a memory, a physical processor in communication with the memory, a first device, a second device, a first hypervisor at a source location, a second hypervisor at a destination location, a first virtual machine, and a second virtual machine. The first virtual machine includes a guest OS executing on the first hypervisor and a guest driver and the second virtual machine includes the guest OS executing on the second hypervisor and the guest driver. The first and second hypervisors send a request to save a device state to the guest driver in the guest OS executing on the first hypervisor and send a migration notification to the guest OS executing on the second hypervisor. The guest driver saves a state signature in the memory. The guest driver determines a status of the state signature as one of matching the second device and mismatching the second device.

Abstract: A computer system receives a first instruction identifier identifying a first instruction being executed by a first thread associated with a first virtual machine and receives a second instruction identifier identifying a second instruction being executed by a second thread associated with the first virtual machine. The system produces an execution record comprising the first instruction identifier, a first thread identifier identifying the first thread, a first timestamp reflecting a first real-time clock value associated with executing the first instruction, the second instruction identifier, a second thread identifier identifying the second thread, and a second timestamp reflecting a second real-time clock value associated with executing the second instruction.

Abstract: Certain aspects direct to systems and methods for platform simulation on virtual machine for development projects of a management controller on virtual machines. At least one virtual machine (VM) is provided to simulate a management controller and a host computing device for the management controller. The at least one VM includes: a firmware module for the management controller, configured to receive at least one output signal from the host computing device or from at least one device connected to the host computing device; and a simulator module configured to simulate the host computing device or the at least one device connected to the host computing device. In operation, the simulator module generates the at least one output signal based on configuration data of the host computing device or the device connected to the host computing device, and sends the at least one output signal to the firmware module.

Abstract: Certain aspects direct to systems and methods for cloud based platform simulation on virtual machine for development projects of a management controller. The system includes a cloud network, which include necessary computing resources to provide at least one virtual machine (VM), such that the at least one VM is accessible on the cloud network to simulate a management controller and a host computing device for the management controller. The at least one VM includes: a firmware module for the management controller, configured to receive at least one output signal from the host computing device; and a simulator module configured to simulate the host computing device. In operation, the simulator module generates the at least one output signal based on configuration data of the host computing device, and sends the at least one output signal to the firmware module.

Abstract: Apparatuses, methods, systems, and computer program products are disclosed for container synchronization. Host data associated with a first container on a host device is determined. Target data associated with a second container on a target device is determined. Differences between the host data associated with the first container on the host device and the target data associated with the second container on the target device are determined. Host data associated with the first container on the host device is synchronized with target data associated with the second container on the target device in response to determining differences between the host data and the target data such that the host data associated with the first container is identical to the target data associated with the second container.

Abstract: The purpose of the present invention is to provide a thin client system using a virtual machine which more satisfactorily suppresses occurrences of login storms. A connection management device 10 comprises: a virtual machine selection unit 11 which, on the basis of usage information of virtual machines 21 and valid profile location information of each user, selects the virtual machines 21 which are to be allocated to the users who log in to a thin client system 1; and a user session management unit 12 which, according to connection states of the users with respect to the virtual machines 21, updates the usage information and the valid profile location information.

Abstract: A storage medium stores a program that causes a computer to execute a process. The process includes, obtaining information on a virtual machine from among a plurality of virtual machines, migration information, first state information, and time information, determining a priority of each of the plurality of virtual machines based on the time information, identifying a first virtual machine group based on the migration information and the first state information, instructing each virtual machine included in the first virtual machine group to migrate to a physical machine of a migration destination based on the priority, obtaining second state information, identifying a second virtual machine group based on the migration information and the second state information, and instructing each virtual machine included in the second virtual machine group to migrate to the physical machine of the migration destination based on the priority.

Abstract: A moving control method performed by a computer, includes: generating a first-conditional expression for a relationship between a first variable for a physical machine on which each of the virtual machines is disposed and a second variable for a movement status of each of the virtual machines, based on the first variable, and the second variable; generating a second-conditional expression for a relationship between the second variable and a third variable, based on the second variable, and a third variable for time taken to move the virtual machines; calculating values of the first variable, the second variable, and the third variable where the third variable is smallest, based on the first-conditional expression, the second-conditional expression; and performing an instruction to move each of the virtual machines, based on the values of the first variable, the second variable, and the third variable where the third variable is smallest.

Abstract: An information processing system includes first data center and second data center. The first data center includes first virtual machine server including first processor that executes virtual machine using an image file, first data server including second processor and first storage that stores a plurality of chunks, and first metadata server including third processor and second storage that stores metadata of the image file. The second data center includes a second virtual machine server including fourth processor, second data server including third storage and fifth processor, and second metadata server including sixth processor and fourth storage. The second processor transmits, to the second data center, a predetermined chunk of the plurality of chunks. The first processor stops the virtual machine after the predetermined chunk is transmitted. The third processor transmits the metadata to the second metadata server. The fourth processor activates the virtual machine.

Abstract: A method, system and computer program product for selecting a target hypervisor to run a migrated virtual machine. An “effective priority value,” representing the virtual machine's priority with respect to the other virtual machines running on the same hypervisor, is calculated for the virtual machine when it is running on the source hypervisor as well as if it were to run on a target hypervisor for each possible target hypervisor. The target hypervisor associated with the minimum difference in absolute value terms between the virtual machine's effective priority value calculated when it is running on the source hypervisor and its effective priority value calculated if it were to be migrated to run on a target hypervisor is selected to receive the migrating virtual machine. In this manner, the effective priority metric has enabled a target hypervisor to be chosen that most closely matches the priority environment of the source hypervisor.

Abstract: In one embodiment, a computer program product includes a computer readable storage medium having program instructions embodied therewith. The embodied program instructions are executable by a processor to cause the processor to receive, by the processor, a first job request. The embodied program instructions are also executable by the processor to cause the processor to analyze, by the processor, the first job request to determine a user skill level of a user that submitted the first job request. Moreover, the embodied program instructions are executable by the processor to cause the processor to admit, by the processor, the first job request to a data analytics system and/or a data storage system in a specified order with respect to other received job requests based on at least the user skill level of the user that submitted the first job request. Other systems and methods are described in accordance with more embodiments.

Abstract: The present disclosure relates to extending object-schema-based application programming interfaces (APIs). According to one embodiment, a method generally includes receiving, from a user, a schema defining an extension to the API and a reference to a parent node in a graph projection of the API. An API system updates the graph projection of the API to include a node representing the extension and navigable path to the node representing the extension. The API system processes a request from the user by traversing through the updated graph projection of the API, the request representing the navigable path to the node representing the extension.

Abstract: Apparatuses, methods, systems, and program products are disclosed for interrupting a device. A method includes detecting, by use of a processor, an interrupt cue in response to input received from the one or more sensors. The method further includes determining one or more actively executing applications on an information handling device. The method also includes interrupting the one or more actively executing applications in response to the interrupt cue.

Abstract: An information processing device includes a confirmation unit and an application unit. If a changeable flag is attached to a process flow, the confirmation unit confirms whether or not a change exists in a master of the process flow. If the confirmation unit confirms that a change exists in the master, the application unit applies the change to the process flow that is to be executed.

Abstract: An application profiling system, initiating profiling a software application; including: apparatus to receive user input information of a software application profiling target and execution requirements, to store profiler specifications; to determine which profiler satisfies the execution requirements, based on the specifications, and to generate needed profiling tasks, each task specifying an application profiler; to select hardware resources the tasks; and to initiate execution of the tasks.

Abstract: A method and an electronic device are provided in which, in response to a first user input, a stack of partially overlaid visual elements is displayed in response to the first user input. Each visual element corresponds to an application that is running in the electronic device and includes an index item representing the corresponding application. A second user input for selecting a visual element from the stack of partially overlaid visual elements is received through the touchscreen. An execution screen of an application corresponding to the selected visual element is displayed.

Abstract: A sequence of tuples, each having one or more attributes, is received at one of one or more processing elements operating on one or more processors. Each processing element may have one or more stream operators. A first stream operator may be identified as one that only processes an instance of a first attribute in a currently received tuple when a difference between an instance of the first attribute in a previously received tuple and the instance of the first attribute in the currently received tuple is outside of a difference threshold. A second stream operator may generate a difference attribute from a first instance of the first attribute in a first one of the received tuples and a second instance of the first attribute in a second one of the received tuples. The difference attribute may be transmitted from the second stream operator to the first stream operator.

Abstract: A sequence of tuples, each having one or more attributes, is received at one of one or more processing elements operating on one or more processors. Each processing element may have one or more stream operators. A first stream operator may be identified as one that only processes an instance of a first attribute in a currently received tuple when a difference between an instance of the first attribute in a previously received tuple and the instance of the first attribute in the currently received tuple is outside of a difference threshold. A second stream operator may generate a difference attribute from a first instance of the first attribute in a first one of the received tuples and a second instance of the first attribute in a second one of the received tuples. The difference attribute may be transmitted from the second stream operator to the first stream operator.

Abstract: A migration system includes a memory, a physical processor, first and second hypervisors, first and second virtual machines, and first and second networking devices. The first hypervisor is located at a migration source location and the second hypervisor is located at a migration destination location. The first virtual machine includes a guest OS which includes a first agent. The second virtual machine includes the guest OS which includes a second agent. The first hypervisor is configured to request the guest OS executing on the first hypervisor to copy a configuration of the first networking device and to store the configuration in a place-holder networking device. The second hypervisor is configured to start the second virtual machine at a destination location, request the guest OS executing on the second virtual machine to copy the configuration from the place-holder networking device and to store the configuration in the second networking device.

Abstract: Methods, apparatuses and storage medium associated with migration between processors by a computing device are disclosed. In various embodiments, a portable electronic device having an internal processor and internal memory may be attached to a dock. The dock may include another processor as well other memory. The attachment of the dock to the portable electronic device may cause an interrupt. In response to this interrupt, a state associated with the internal processor may be copied to the other memory of the dock. Instructions for the computing device may then be executed using the other processor of the dock. Other embodiments may be disclosed or claimed.

Abstract: Data-locality-aware task scheduling on hyper-converged computing infrastructures is provided. A plurality of data blocks referenced in an input/output (I/O) request are identified. The I/O request is based on scheduling logic that executes within a container that is deployed on a hyper-converged infrastructure. A block-location mapping table is scanned using a data block identifier that is associated with a present data block of the plurality of data blocks. Physical node(s) of the hyper-converged infrastructure that store the present data block are identified. A container-instance mapping table is scanned using one or more physical node identifiers that are associated with the physical node(s) that store the present data block. Container(s) deployed on physical node(s) that store the present data block are identified. The scheduling logic is provided with a list of identifier(s) that are respectively associated with the containers that are deployed on the physical node(s) that store the present data block.

Abstract: A control apparatus is communicably connected to a plurality of processing apparatuses, including a processor configured to determine whether the sum of an execution time of a first process, an execution time of a second process, and a time taken for a first processing apparatus among the plurality of processing apparatuses to rewrite a logic for executing the first process to a logic for executing the second process is equal to or smaller than a unit time; determine whether a data traffic between the plurality of processing apparatuses is equal to or smaller than a threshold when the first processing apparatus executes the first and second processes, and cause the first processing apparatus to execute the first and second processes when it is determined that the sum is equal to or smaller than the unit time and the data traffic is equal to or smaller than the threshold.

Abstract: A parallel processing apparatus includes a plurality of calculation processors configured to execute a job, a memory configured to store information of an executed job before a target job is submitted, an execution exit code of the executed job, the target job, a submitted job before the target job is submitted, and a time difference between a timing of an immediately preceding event of the target job and a timing of submitting the target job, and a management processor coupled to the memory and configured to predict a period to a timing of submitting a next job and a necessity number of the calculation processors for calculating the next job by machine learning mechanism on the basis of the information stored in the memory when an event occurs, and control each of the calculation processors in accordance with the predicted period and the necessity number of the calculation processors.

Abstract: A plurality of processing entities are maintained in a processor complex. In response to determining that a task is a critical task, the critical task is dispatched to a scheduler, wherein it is preferable to prioritize execution of critical tasks over non-critical tasks. In response to dispatching the critical task to the scheduler, the scheduler determines which processing entity of the plurality of processing entities has a least amount of processing remaining to be performed for currently scheduled tasks. Tasks queued on the determined processing entity are moved to other processing entities, and the currently scheduled tasks on the determined processing entity are completed. In response to moving tasks queued on the determined processing entity to other processing entities and completing the currently scheduled tasks on the determined processing entity, the critical task is dispatched on the determined processing entity.

Abstract: A plurality of cores are maintained in a processor complex. A core of the plurality of cores is reserved for execution of critical tasks, wherein it is preferable to prioritize execution of critical tasks over non-critical tasks. A scheduler receives a task for scheduling in the plurality of cores. In response to determining that the task is a critical task, the task is scheduled for execution in the reserved core.

Abstract: An apparatus allocating controller is provided in the invention. The apparatus allocating controller includes a bus, a communication interface, and a processor. The bus is coupled to a peripheral apparatus, and the peripheral apparatus meets the peripheral component interconnect express. The communication interface is coupled to a server. The apparatus allocating controller and the peripheral apparatus communicate with the server through the communication interface. The processor is coupled to the bus. The processor is configured to allocate the peripheral apparatus to the server, and monitors the connection status of the peripheral apparatus and the bus. When another peripheral apparatus hot plugs in the bus so that a hot-plug event occurs on the bus, the processor determines whether the another peripheral apparatus has a physical function or a virtual function, and the processor reallocates the peripheral apparatus to the server according to the hot-plug event.

Abstract: A management system includes an interface device coupled to a computer system including a plurality of resources including resources of a plurality of resource types and a processor coupled to the interface device. The processor allocates, to a tenant environment, a resource of a resource type, which is different depending on an XaaS (X as a Service), among the plurality of resources including resources of a plurality of resource types and provides a plurality of tenant environments including a plurality of XaaSs. The processor manages, as low-order resources, resources lower in order than the resources allocated to the tenant environments, among the plurality of resources including resources of a plurality of resource types. When receiving a resource change request, the processor executes change processing on at least one of the low-order resources among the resources managed as low-order resources.

Abstract: Systems and methods for resolving application multitasking degradation are disclosed. In aspects, a computer implemented method is used with a user device including a multitasking operating system, shared user device resources, a first application and a second application. The method includes: running, simultaneously, the first application and the second application; measuring performance parameters for one or more application tasks of the first and second applications; and determining that one or more of the performance parameters of the one or more application tasks falls below a performance threshold value of an associated key performance indicator (KPI). The determination indicates degradation in performance of at least one of the first application and second application. The method further includes instructing the operating system to modify an allocation of the shared user device resources to address the degradation in performance of the at least one of the first application and second application.

Abstract: Techniques for tracking large numbers of moving objects in an event processing system are provided. An input event stream can be received, where the events in the input event stream represent the movement of a plurality of geometries or objects. The input event stream can then be partitioned among a number of processing nodes of the event processing system, thereby enabling parallel processing of one or more continuous queries for tracking the objects. The partitioning can be performed such that each processing node is configured to track objects in a predefined spatial region, and the spatial regions for at least two nodes overlap. This overlapping window enables a single node to find, e.g., all of the objects within a particular distance of a target object, even if the target object is in the process of moving from the region of that node to the overlapping region of another node.

Abstract: A method for managing compute and I/O tasks in a data processing system includes: providing a thread pool including a plurality of thread groups, each thread group including one or more threads; providing a CPU pool including a plurality of CPU groups, each CPU group including one or more CPU cores; receiving a plurality of tasks comprising I/O tasks and compute tasks; mapping each of the plurality of tasks to at least one thread group in the thread pool; and mapping each of the plurality of thread groups in the thread pool to at least one CPU group in the CPU pool. The mappings between the plurality of tasks and the plurality of thread groups and between the plurality of thread groups and the plurality of CPU groups dynamically change based on performance variables.

Abstract: Disaggregated computing architectures, platforms, and systems are provided herein. In one example, a method of operating a disaggregated computing architecture is presented. The method includes, receiving user commands to establish compute units among a plurality of physical computing components, each of the compute units comprising one or more of the plurality of physical computing components selected from among central processing units (CPUs), graphics processing units (GPUs), storage modules, and network interface modules. The method also includes forming the compute units based at least on logical partitioning within a Peripheral Component Interconnect Express (PCIe) fabric communicatively coupling the plurality of physical computing components, wherein each of the compute units have visibility over the PCIe fabric to the one or more of the plurality of physical computing components assigned to the associated compute units using the logical partitioning within the PCIe fabric.