Abstract: A memory access command, column address and plurality of write data values are received within an integrated-circuit memory chip via external signaling links. In response to the memory access command, the integrated-circuit memory chip (i) decodes the column address to select address-specified sense amplifiers from among a plurality of sense amplifiers that constitute a sense amplifier bank, (ii) reads first data, constituted by a plurality of read data values, out of the address-specified sense amplifiers, and (iii) overwrites the first data within the address-specified sense amplifiers with second data constituted by one or more of the write data values and by one or more of the read data values.

Abstract: An intelligent tile-based memory bandwidth management solution executed by an address aperture, such as a compression address aperture, services linearly addressed data requests (read requests and write requests) from a processor to data stored in a memory component having a tile-based address structure. For read requests, the aperture stores previously read tiles (full or partial) in a tile-aware cache and then seeks to service future read requests from the cache instead of the long-term memory component. For write requests, the aperture stores the write data in the tile-aware cache and assembles the data with write data from other write requests so that full tile data writes to the long-term memory may be achieved in lieu of excessive partial-tile writes.

Abstract: An intelligent tile-based prefetching solution executed by a compression address aperture services linearly addressed data requests from a processor to memory stored in a memory component having a tile-based address structure. The aperture monitors tile reads and seeks to match the tile read pattern to a predefined pattern. If a match is determined, the aperture executes a prefetching algorithm uniquely and optimally associated with the predefined tile read pattern. In this way, tile overfetch is mitigated while the latency on first line data reads is reduced.

Abstract: Systems and methods enable displaying a graphical representation of system resource usage in a resource utilization map to inform users about system resource utilization by applications and processes running on a computing device. Users may provide inputs to enable the system to adjust resource allocations based on user preferences. This may enable users to improve the overall operational performance of the device consistent with their current personal preferences by identifying applications or processes of most or least interest so the device processor to prioritize system resources accordingly. Some aspects transmit resource allocation data based on such user input to a central server to enable community based resource allocation schemes. Community based resource allocation schemes may be transmitted to computing devices for use as default or preliminary resource allocations for particular applications, websites or device operating states.

Abstract: A memory access command, column address and plurality of write data values are received within an integrated-circuit memory chip via external signaling links. In response to the memory access command, the integrated-circuit memory chip (i) decodes the column address to select address-specified sense amplifiers from among a plurality of sense amplifiers that constitute a sense amplifier bank, (ii) reads first data, constituted by a plurality of read data values, out of the address-specified sense amplifiers, and (iii) overwrites the first data within the address-specified sense amplifiers with second data constituted by one or more of the write data values and by one or more of the read data values.

Abstract: Systems and methods for managing coherency in a processing system comprising a memory involve one or more aperture cache coherency (ACC) blocks. The ACC blocks monitor accesses to the memory using aliased addresses, wherein the aliased addresses map to locations in an aliased address domain of the memory. The ACC blocks also monitor accesses to the memory through a functional address aperture using aperture addresses, wherein a function of the aperture addresses map to locations in an aperture address domain of the memory. The ACC blocks are further configured to maintain coherency between one or more of data in a first location of the memory, the first location belonging to the aliased address domain and the aperture address domain; one or more copies of the data accessed using the aperture addresses; or one or more copies of the data accessed using the aliased addresses.

Abstract: Methods and systems are disclosed for full-hardware management of power and clock domains related to a distributed virtual memory (DVM) network. An aspect includes transmitting, from a DVM initiator to a DVM network, a DVM operation, broadcasting, by the DVM network to a plurality of DVM targets, the DVM operation, and, based on the DVM operation being broadcasted to the plurality of DVM targets by the DVM network, performing one or more hardware optimizations comprising: turning on a clock domain coupled to the DVM network or a DVM target of the plurality of DVM targets that is a target of the DVM operation, increasing a frequency of the clock domain, turning on a power domain coupled to the DVM target based on the power domain being turned off, or terminating the DVM operation to the DVM target based on the DVM target being turned off.

Abstract: Methods and systems are disclosed for full-hardware management of power and clock domains related to a distributed virtual memory (DVM) network. An aspect includes transmitting, from a DVM initiator to a DVM network, a DVM operation, broadcasting, by the DVM network to a plurality of DVM targets, the DVM operation, and, based on the DVM operation being broadcasted to the plurality of DVM targets by the DVM network, performing one or more hardware optimizations comprising: turning on a clock domain coupled to the DVM network or a DVM target of the plurality of DVM targets that is a target of the DVM operation, increasing a frequency of the clock domain, turning on a power domain coupled to the DVM target based on the power domain being turned off, or terminating the DVM operation to the DVM target based on the DVM target being turned off.

Abstract: Aspects include computing devices and methods implemented by the computing for automatic cache coherency for page table data on a computing device. Some aspects may include modifying, by a first processing device, page table data stored in a first cache associated with the first processing device, receiving, at a page table coherency unit, a page table cache invalidate signal from the first processing device, issuing, by the page table coherency unit, a cache maintenance operation command to the first processing device, and writing, by the first processing device, the modified page table data stored in the first cache to a shared memory accessible by the first processing device and a second processing device associated with a second cache storing the page table data.

Abstract: This disclosure relates to allocating memory resources of a computing device comprising non-volatile random access memory (NVRAM) and dynamic random access memory (DRAM). An exemplary method is performed for every independently executable component of an application and includes determining attributes of the component. The method also includes associating the component with a memory profile of a plurality of memory profiles based on the attributes, wherein each memory profile of the plurality of memory profiles specifies a number of banks of the NVRAM and a number of banks of the DRAM. The method also includes causing the computing device to generate an assignment of the component to banks of the NVRAM and DRAM based on the memory profile associated with the component so the computing device can execute the component using the banks of the NVRAM and DRAM based on the assignment.

Abstract: Methods, devices, and non-transitory process-readable storage media for compacting data within cache lines of a cache. An aspect method may include identifying, by a processor of the computing device, a base address (e.g., a physical or virtual cache address) for a first data segment, identifying a data size (e.g., based on a compression ratio) for the first data segment, obtaining a base offset based on the identified data size and the base address of the first data segment, and calculating an offset address by offsetting the base address with the obtained base offset, wherein the calculated offset address is associated with a second data segment. In some aspects, the method may include identifying a parity value for the first data segment based on the base address and obtaining the base offset by performing a lookup on a stored table using the identified data size and identified parity value.

Abstract: Aspects include computing devices, apparatus, and methods implemented by the apparatus for implementing dynamic input/output (I/O) coherent workload processing on a computing device. Aspect methods may include offloading, by a processing device, a workload to a hardware accelerator for execution using an I/O coherent mode, detecting a dynamic trigger for switching from the I/O coherent mode to a non-I/O coherent mode while the workload is executed by the hardware accelerator, and switching from the I/O coherent mode to a non-I/O coherent mode while the workload is executed by the hardware accelerator.

Abstract: Aspects include computing devices, apparatus, and methods implemented by the apparatus for implementing a hybrid input/output (I/O) coherent write request on a computing device, including receiving an I/O coherent write request, generating a first hybrid I/O coherent write request and a second hybrid I/O coherent write request from the I/O coherent write request, sending the first hybrid I/O coherent write request and I/O coherent write data of the I/O coherent write request to a shared memory, and sending the second hybrid I/O coherent write request without the I/O coherent write data of the I/O coherent write request to a coherency domain.

Abstract: Various aspects include methods and computing devices implementing methods for reverse tiling of work items. Various aspects may include receiving information relating to a kernel execution, receiving information relating to a work item created for a kernel execution, and applying a reverse tiling function to produce a reverse tiling work item identifier (ID) for the work item to implement a pattern of access of the memory device resources. In various aspects, the reverse tiling function may be a static preprogrammed reverse tiling function, a a dynamically generated reverse tiling function, or a reverse tiling function selected from a plurality of reverse tiling functions. In various aspects, applying the reverse tiling function to produce the reverse tiling work item identifier for the work item may occur in response to determining that the pattern of access of a memory device resources provides a benefit over a default pattern of access.

Abstract: Aspects include computing devices and methods implemented by the computing for automatic cache coherency for page table data on a computing device. Some aspects may include modifying, by a first processing device, page table data stored in a first cache associated with the first processing device, receiving, at a page table coherency unit, a page table cache invalidate signal from the first processing device, issuing, by the page table coherency unit, a cache maintenance operation command to the first processing device, and writing, by the first processing device, the modified page table data stored in the first cache to a shared memory accessible by the first processing device and a second processing device associated with a second cache storing the page table data.

Abstract: Aspects include computing devices, apparatus, and methods implemented by the apparatus for input/output-coherent look-ahead cache access on a computing device.

Abstract: Various aspects are described herein. In some aspects, the present disclosure provides a method of communicating data between an electronic unit of a system-on-chip (SoC) and a dynamic random access memory (DRAM). The method includes initiating a memory transaction corresponding to first data. The method includes determining a non-unique first signature and a unique second signature associated with the first data based on content of the first data. The method includes determining if the non-unique first signature is stored in at least one of a local buffer on the SoC separate from the DRAM or the DRAM. The method includes determining if the unique second signature is stored in at least one of the local buffer or the DRAM based on determining the non-unique first signature is stored. The method includes eliminating the memory transaction with respect to the DRAM based on determining the unique second signature is stored.

Abstract: A memory access command, column address and plurality of write data values are received within an integrated-circuit memory chip via external signaling links. In response to the memory access command, the integrated-circuit memory chip (i) decodes the column address to select address-specified sense amplifiers from among a plurality of sense amplifiers that constitute a sense amplifier bank, (ii) reads first data, constituted by a plurality of read data values, out of the address-specified sense amplifiers, and (iii) overwrites the first data within the address-specified sense amplifiers with second data constituted by one or more of the write data values and by one or more of the read data values.

Abstract: Various aspects are described herein. In some aspects, the present disclosure provides a method of communicating data between an electronic unit of a system-on-chip (SoC) and a dynamic random access memory (DRAM). The method includes initiating a memory transaction corresponding to first data. The method includes determining a non-unique first signature and a unique second signature associated with the first data based on content of the first data. The method includes determining if the non-unique first signature is stored in at least one of a local buffer on the SoC separate from the DRAM or the DRAM. The method includes determining if the unique second signature is stored in at least one of the local buffer or the DRAM based on determining the non-unique first signature is stored. The method includes eliminating the memory transaction with respect to the DRAM based on determining the unique second signature is stored.

Abstract: Aspects include computing devices, apparatus, and methods implemented by the apparatus for implementing a hybrid input/output (I/O) coherent write request on a computing device, including receiving an I/O coherent write request, generating a first hybrid I/O coherent write request and a second hybrid I/O coherent write request from the I/O coherent write request, sending the first hybrid I/O coherent write request and I/O coherent write data of the I/O coherent write request to a shared memory, and sending the second hybrid I/O coherent write request without the I/O coherent write data of the I/O coherent write request to a coherency domain.