A device in an interconnect network is provided. The device comprises an end point processor comprising end point memory and an interconnect network link in communication with an interconnect network switch. The device is configured to issue, by the end point processor, a request to send data from the end point memory to other end point memory of another end point processor of another device in the interconnect network and provide, to the interconnect network switch, the request using memory addresses from a global memory address map which comprises a first global memory address range for the end point processor and a second global memory address range for the other end point processor.

The present disclosure relates to asymmetric read/write architectures for enhanced throughput and reduced latency. One example embodiment includes an integrated circuit. The integrated circuit includes a network interface. The integrated circuit also includes a communication bus interface. The integrated circuit is configured to establish a communication link with a processor of the host computing device over the communication bus interface, which includes mapping to memory addresses associated with the processor of the host computing device. The integrated circuit is also configured to receive payload data for transmission over the network interface in response to the processor of the host computing device writing payload data to the mapped memory addresses using one or more programmed input-outputs (PIOs). Further, the integrated circuit is configured to write payload data received over the network interface to the memory of the host computing device using direct memory access (DMA).

Circuits and methods enabling common control of an agent device by two or more buses, particularly MIPI RFFE serial buses. In essence, the invention provides flagging signals designating completed register write operations to denote which of two registers are active, such that synchronization is accomplished in a clock-free manner. One embodiment includes at least two decoders, each including a common register and a bus (S/P) decoder coupled to a respective bus and to the common register. The S/P decoder asserts a write-complete signal when a write operation to a corresponding common register is completed. A multiplexer has at least two selectable input bus ports coupled to the common registers within the at least two decoders. A selection circuit selects an input bus port of the multiplexer in response to the assertion of a last write-complete signal from the S/P decoders.

A semiconductor device capable of shortening a time required for data transfer and data organizing is provided. The solid state device includes a processor, a memory, an external interface, registers for storing data received by the external interface, a mirror register buffer, a processor, a memory, an external interface, registers, and an internal bus connected to the mirror register buffer. Registers output data to the mirror register buffer without going through the internal bus. Mirror register buffer gives the data input from the registers an address in a mirror register buffer different from the address allocated to the register, and transfers the data to the memory without passing through the internal bus.

The present disclosure relates to systems, methods, and computer-readable media for data from a first multi-dimensional memory block to a second multi-dimensional memory block. For example, systems described herein facilitate transferring data between memory blocks having different shapes from one another. The systems described herein facilitate transferring data between different shaped memory blocks by identifying shape properties and other characteristics of the data and generating a plurality of network packets having control data based on the identified shape properties and other characteristics. This data included within the network packets enables memory controllers to determine memory addresses on a destination memory block to write data from the network packets. Features described herein facilitate efficient transfer of data without generating a linearized copy that relies on constant availability of significant memory resources.

Channel availability information associated with data traffic between a Master and a Slave within an interconnection network (“ICN”) in a System-on-Chip (“SoC”) is monitored by a channel performance monitor in order to improve the performance of the ICN. The channel availability information is fed back to certain Masters to control their data traffic into the ICN. The channel performance monitor monitors and evaluates the data traffic handled by switches within the ICN that can potentially interfere with communication paths between particular Masters and Slaves, and control the initiation of data traffic from predetermined Masters.

An electronic device comprises a module configured to transfer data bus transactions from a transaction source domain to a transaction target domain. A first interface receives the transaction from the source domain using a transaction source ID. A second interface sends the transaction to the target domain using a transaction target ID. A look-up table has a plurality of index values and stores the transaction source ID against one of the index values. Mapping logic determines whether the look-up table contains the transaction source ID stored against any of the index values. When the transaction source ID is already stored, the transaction target ID is set to that index value. Conversely, when the transaction source ID is not stored, an available index value is selected, the transaction source ID is stored against that available index value, and the transaction target ID is set to that available index value.

A shared memory controller receives, from a computing node, a request associated with a memory transaction involving a particular line in a memory pool. The request includes a node address according to an address map of the computing node. An address translation structure is used to translate the first address into a corresponding second address according to a global address map for the memory pool, and the shared memory controller determines that a particular one of a plurality of shared memory controllers is associated with the second address in the global address map and causes the particular shared memory controller to handle the request.

Methods and apparatus for efficient scaling of fabric architectures such as those based on PCIe technology, including up to very large fabrics and numbers of hosts/devices for use in ultra-high performance applications such as for example data centers and computing clusters. In one aspect, methods and apparatus for using Non-Transparent Bridge (NTB) technology to export Message Signaled Interrupts (MSIs) to external hosts are described. In a further aspect, an IO Virtual Address (IOVA) space is created is used as a method of sharing an address space between hosts, including across the foregoing NTB(s). Additionally, a Fabric Manager (FM) entity is disclosed and utilized for programming e.g., PCIe switch hardware to effect a desired host/fabric configuration.

A dynamic loading system of an off-chip non-volatile memory based on virtual mapping includes the off-chip non-volatile memory, an on-chip memory, an on-chip mapping device, a memory access controller, and an off-chip memory. The on-chip mapping device is configured to construct a mapping relationship between a logical partition and a physical partition of the on-chip memory. The memory access controller is configured to parse a memory access of a system into an access to the logical partition, configure a prediction algorithm of data loading, and send a memory loading request to the off-chip memory loading device. The off-chip memory loading device is configured to construct a mapping relationship between a program partition of the off-chip non-volatile memory and the physical partition of the on-chip memory.