Systems and methods are directed to managing access to a shared memory. A request received at a memory controller, for access to the shared memory from a client of one or more clients configured to access the shared memory, is placed in at least one queue in the memory controller. A series of one or more timeout values is assigned to the request, based, at least in part on a priority associated with the client which generated the request. The priority may be fixed or based on a Quality-of-Service (QoS) class of the client. A timer is incremented while the request remains in the first queue. As the timer traverses each one of the one or more timeout values in the series, a criticality level of the request is incremented. A request with a higher criticality level may be prioritized for servicing over a request with a lower criticality level.

A system, methods, and apparatus for determining memory distribution across multiple non-uniform memory access processing nodes are disclosed. An apparatus includes processing nodes, each including processing units and main memory serving as local memory. A bus connects the processing units of each processing node to different main memory of a different processing node as shared memory. Access to local memory has lower memory access latency than access to shared memory. The processing nodes execute threads distributed across the processing nodes, and detect memory accesses made from each processing node for each thread. The processing nodes determine locality values for the thread that represent the fraction of memory accesses made from the processing nodes, and determine processing time values for the threads for a sampling period. The processing nodes determine weighted locality values for the threads, and determine a memory distribution across the processing nodes based on the weighted locality values.

A data processing architecture includes a processor which may access a memory and fetch a command recorded in the memory, transmit the fetched command to a subject configured to perform an operation corresponding to the fetched command through a network, and receive a result of performing the operation from the subject and record the result of performing the operation in the memory.

Methods and systems for mutual exclusion in a non-coherent memory hierarchy may include a non-coherent memory system with a shared system memory. Multiple processors and a memory connect interface may be configured to provide an interface for the processors to the shared memory. The memory connect interface may include an arbiter for atomic memory operations from the processors. In response to an atomic memory operation, the arbiter may perform an atomic memory operation procedure including setting a busy flag for an address of the atomic memory operation, blocking subsequent memory operations from any of the processors to the address while the busy flag is set, issuing the atomic memory operation to the shared memory, and in response to an acknowledgement of the atomic memory operation from the shared memory, clearing the busy flag and allowing subsequent memory operations from the processors for the address to proceed to the shared memory.

The systems and methods are configured to efficiently and effectively access memory. In one embodiment, a memory controller comprises a request queue, a buffer, a control component, and a data path system. The request queue receives memory access requests. The control component is configured to process information associated with access requests via a first narrow memory channel and a second narrow memory channel. The first narrow memory channel and the second narrow memory channel can have a portion of command/control communication lines and address communication lines that are included in and shared between the first narrow memory channel and the second narrow memory channel. The data path system can include a first data module and one set of unshared data lines associated with the first memory channel and a second data module and another set of unshared data lines associated with second memory channel.

A guest OS detects a DMA write request for a device assigned to the guest OS to perform a DMA write to a shared page of memory that has a write protection attribute to cause a protection page fault upon an attempt to write to the shared page of memory. The guest OS reads a portion of the shared page of memory from a location of that page, determines the value of the portion, and executes an atomic instruction that writes the value back to the location of the shared page of memory to trigger the page protection fault. Upon executing the atomic instruction, the guest OS sends the DMA write request to the device to cause the device to write to a writeable copy of the shared page of memory.

A memory controller and method are provided for controlling a memory device to process access requests issued by at least one master device, the memory device having a plurality of access regions. The memory controller has a pending access requests storage that buffers access requests that have been issued by a master device prior to those access requests being processed by the memory device. Access control circuitry then issues control commands to the plurality of access regions in order to control the memory device to process access requests retrieved from the pending access requests storage. A query structure is also provided that is configured to maintain, for each access region, information about the buffered access requests in the pending access requests storage, and the access control circuitry references the query structure when determining the control commands to be issued to the plurality of access regions. Such an approach enables significant performance and energy savings to be realized in control of the memory device, without requiring the contents of the pending access requests storage to be directly monitored by the access control circuitry.

A computer system may support one or more techniques to allow dynamic pinning of the memory pages accessed by a non-CPU device, such as a graphics processing unit (GPU). The non-CPU may support virtual to physical address mapping and may thus be aware of the memory pages, which may not be pinned but may be accessed by the non-CPU. The non-CPU may notify or send such information to a run-time component such as a device driver associated with the CPU. The device driver may, dynamically, perform pinning of such memory pages, which may be accessed by the non-CPU. The device driver may even unpin the memory pages, which may be no longer accessed by the non-CPU. Such an approach may allow the memory pages, which may be no longer accessed by the non-CPU to be available for allocation to the other CPUs and/or non-CPUs.

A microcontroller includes two processing blocks that respectively have a Central Processing Unit (CPU) and a peripheral circuit, where an access to the peripheral circuit in each of the processing blocks, that is, to a Read-Only Memory (ROM) or a Pulse Width Modulator (PWM) signal generator, is limited only from the CPU disposed in the same processing block. Thereby a fail-safe functionality of the microcontroller is improved.

According to an embodiment, an information processing apparatus includes a plurality of cores, a shared resource that can be shared by the plurality of cores, and local registers that store configuration information peculiar to the respective cores. The shared resource is provided independently from the plurality of cores. The local registers are provided to the respective cores. This makes it possible to provide an information processing apparatus that can suppress increase in hardware resources even when the number of cores composing a multi-core system increases.