Abstract: The present application provides a data backup method and a restoration method for an NVDIMM, an NVDIMM controller and an NVDIMM. The NVDIMM (200) comprises a DRAM (201), a NAND flash memory (202) and an NVDIMM controller (100), the NVDIMM controller (100) controlling the NVDIMM (200) and comprising a DDR controller (101), a NAND flash memory controller (102), a data backup module (103) and a data restoration module (104), the DDR controller (101) using and enabling DBI mechanism. During data backup, the DDR controller (101) reads N-bit DQi and 1-bit DBI from the DRAM (201) and sends the same to the data backup module (103). When DBIi is “1”, the data backup module (103) compares the DQi and DQi-1. If the number of bits of the DQi and the DQi-1 with different values is greater than N/2, then the DQi is inverted and the DBIi is set to “0”, and otherwise the DQi and the DBIi are remained unchanged. When the DBIi is “0”, the DQi and the DBIi are remained unchanged.

Abstract: An LLC chip and a cache system are provided. The LLC chip includes a storage wafer, an interface logic unit and a packaging substrate. The interface logic unit and the storage wafer are arranged in sequence on the packaging substrate. A plurality of processing assemblies is connected to the interface logic unit so as to perform read and write operations on the storage wafer through the interface logic unit. The storage wafer includes at least one storage space. The plurality of processing assemblies performs read and write operations on a specific storage space or any storage space so as to achieve non-shared independent storage access or shared storage access. Signal transmission bandwidth is thus improved by means of the plurality of distributed interfaces, and data is cached by means of a non-shared mode or shared mode so as to increase data accessing efficiency of the processing assemblies.

Abstract: The NVDIMM (200) comprises a DRAM (201), a NAND flash memory (202) and a NVDIMM controller (100), the NVDIMM controller (100) controlling the NVDIMM (200) and comprising a DDR controller (101), a NAND flash memory controller (102), a data backup module (103) and a data recovery module (104), and the DDR controller (101) using and enabling DBI. The backup method comprises: reading, by the DDR controller (101), N-bit DQ and 1-bit DBI from the DRAM (201) and sending the same to the data backup module; encoding, by the data backup module (103), the N-bit DQ and the 1-bit DBI into N-bit EDQ according to the values of the N-bit DQ and the 1-bit DBI, and sending the N-bit EDQ to the NAND flash memory controller; and receiving, by the NAND flash memory controller (102), the N-bit EDQ and writing the N-bit EDQ into the NAND flash memory (202).

Abstract: A graphflow apparatus includes an information buffer (IB) and a load queue (LQ). The IB is configured to cache an instruction queue. The LQ is used to cache a read instruction queue. The IB includes a speculative bit and a speculative identity (ID) field. The speculative bit indicates whether a current instruction is a speculatively-executable instruction. The speculative ID field stores a speculative ID of one speculative operation on the current instruction.

Abstract: This application provides a graph instruction processing method and apparatus. The method is applied to a processor, and includes: detecting whether a first input and a second input of a first graph instruction are in a ready-to-complete state, where the first input and/or the second input are or is a dynamic data input or dynamic data inputs of the first graph instruction.

Abstract: Embodiments of this application disclose apparatuses, processing methods, and related devices An example apparatus includes at least one processing engine (PE), and each of the at least one PE includes M status buffers, an arbitration logic circuit, and X operation circuits. Each of the M status buffers is configured to store status data of one iterative computing task. The arbitration logic circuit is configured to determine, based on the status data in the each of the M status buffers, L graph computing instructions to be executed in a current clock cycle, and allocate the L graph computing instructions to the X operation circuits. Each of the X operation-units circuits is configured to execute a graph computing instruction allocated by the arbitration logic circuit.

Abstract: Disclosed are a graph instruction processing method and apparatus, which relates to the field of computer technologies One example method includes: detecting whether a first graph instruction has a conditional instruction element; and when the first graph instruction has the conditional instruction element, determining that the first graph instruction is a conditional execution instruction, and processing the first graph instruction when both data flow information and control flow information of the first graph instruction are in a ready state; or when the first graph instruction does not have a conditional instruction element, determining that the first graph instruction is a non-conditional execution instruction, and processing the first graph instruction when data flow information of the first graph instruction is in a ready state.

Abstract: This application discloses a processor, a processing method, and a related device. The processor includes a processor core. The processor core includes an instruction dispatching unit and a graph flow unit and at least one general-purpose operation unit that are connected to the instruction dispatching unit. The instruction dispatching unit is configured to: allocate a general-purpose calculation instruction in a decoded to-be-executed instruction to the at least one general-purpose calculation unit, and allocate a graph calculation control instruction in the decoded to-be-executed instruction to the graph calculation unit, where the general-purpose calculation instruction is used to instruct to execute a general-purpose calculation task, and the graph calculation control instruction is used to instruct to execute a graph calculation task. The at least one general-purpose operation unit is configured to execute the general-purpose calculation instruction.

Abstract: Embodiments of the present invention provide a method and apparatuses for monitoring a system bus. The method includes: when a monitoring apparatus monitors a system bus corresponding to the monitoring apparatus and detects that an exception occurs in the system bus, receiving the exception that occurs in the system bus and an identifier of the system bus that are sent by the monitoring apparatus; if the exception that occurs in the system bus is a timeout exception, sending a switch message to a device that is connected to the system bus, so that the device sends and receives a command or data by using a backup system bus; and if not, increasing the number of exception times of the system bus; and when the number of the exception times exceeds a preset threshold, using the backup system bus to transmit data or command.

Abstract: Embodiments of the present invention provide a method and apparatuses for monitoring a system bus. The method includes: performing, by a monitoring apparatus, real-time monitoring on a corresponding system bus, and when detecting that a command is transmitted through the system bus, obtaining command information; determining, according to the command information, whether a command transmission exception occurs in the system bus; if no command transmission exception occurs in the system bus, when detecting that data corresponding to the command is transmitted through the system bus, determining, according to the data and the amount of the data, whether a command exception occurs in the system bus; and when detecting that a response message corresponding to the command is transmitted through the system bus, obtaining a command wait time of the command, and determining, according to the command wait time, whether a timeout exception occurs in the system bus.

Abstract: A method and an apparatus for implementing multi-processor memory coherency are disclosed. The method includes: a Level-2 (L2) cache of a first cluster receives a control signal of the first cluster for reading first data; the L2 cache of the first cluster reads the first data in a Level-1 (L1) cache of a second cluster through an Accelerator Coherency Port (ACP) of the L1 cache of the second cluster if the first data is currently maintained by the second cluster, where the L2 cache of the first cluster is connected to the ACP of the L1 cache of the second cluster; and the L2 cache of the first cluster provides the first data read to the first cluster for processing. The technical solution under the present invention implements memory coherency between clusters in the ARM Cortex-A9 architecture.

Abstract: A method and an apparatus for implementing multi-processor memory coherency are disclosed. The method includes: a Level-2 (L2) cache of a first cluster receives a control signal of the first cluster for reading first data; the L2 cache of the first cluster reads the first data in a Level-1 (L1) cache of a second cluster through an Accelerator Coherency Port (ACP) of the L1 cache of the second cluster if the first data is currently maintained by the second cluster, where the L2 cache of the first cluster is connected to the ACP of the L1 cache of the second cluster; and the L2 cache of the first cluster provides the first data read to the first cluster for processing. The technical solution under the present invention implements memory coherency between clusters in the ARM Cortex-A9 architecture.