WO2021244168A1 - Système sur puce, procédé de transmission de données et modules de diffusion - Google Patents

Système sur puce, procédé de transmission de données et modules de diffusion Download PDF

Info

Publication number
WO2021244168A1
WO2021244168A1 PCT/CN2021/088442 CN2021088442W WO2021244168A1 WO 2021244168 A1 WO2021244168 A1 WO 2021244168A1 CN 2021088442 W CN2021088442 W CN 2021088442W WO 2021244168 A1 WO2021244168 A1 WO 2021244168A1
Authority
WO
WIPO (PCT)
Prior art keywords
module
broadcast
node module
data
slave
Prior art date
Application number
PCT/CN2021/088442
Other languages
English (en)
Chinese (zh)
Inventor
李炜
曹庆新
Original Assignee
深圳云天励飞技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳云天励飞技术股份有限公司 filed Critical 深圳云天励飞技术股份有限公司
Publication of WO2021244168A1 publication Critical patent/WO2021244168A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/76—Architectures of general purpose stored program computers
    • G06F15/78—Architectures of general purpose stored program computers comprising a single central processing unit
    • G06F15/7807—System on chip, i.e. computer system on a single chip; System in package, i.e. computer system on one or more chips in a single package
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/76—Architectures of general purpose stored program computers
    • G06F15/78—Architectures of general purpose stored program computers comprising a single central processing unit
    • G06F15/7807—System on chip, i.e. computer system on a single chip; System in package, i.e. computer system on one or more chips in a single package
    • G06F15/781—On-chip cache; Off-chip memory
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/54—Interprogram communication
    • G06F9/542—Event management; Broadcasting; Multicasting; Notifications
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Definitions

  • This application belongs to the field of electronic technology, and in particular relates to a system on a chip, a data transmission method, and a broadcast module.
  • SoC System-on-a-chip
  • Master master devices
  • Slave slave devices
  • Each master device and each slave device are connected through a bus to realize data Transmit.
  • the bus can only be occupied by one master device and one slave device at any time. If multiple master devices need to request data from the same slave device, the multiple master devices can only press The sequence waits to obtain the right to use the bus, and reads data from the slave device in multiple times.
  • the embodiments of the present application provide a system-on-chip, a data transmission method, and a broadcast module to solve the problems of how to improve the data transmission efficiency of the system-on-chip and reduce power consumption in the prior art.
  • the first aspect of the embodiments of the present application provides a system on chip, the system on chip includes at least two master devices, at least one slave device, and a bus; the bus includes at least two broadcast modules, and each of the broadcast modules Correspondingly connected to one of the main devices; wherein, the broadcast module configured as the initial node in the broadcast topology is the initial node module, and the broadcast modules other than the initial node module in the broadcast topology are Slave node module;
  • the main device is configured to send a data read request and obtain target data through the broadcast module;
  • the initial node module, and each of the broadcast modules is connected to one of the master devices; wherein, the broadcast module configured as the initial node in the broadcast topology is the initial node module, and in the broadcast topology
  • the broadcast module other than the initial node module is a slave node module;
  • the slave node module is configured to receive a second data read request sent by a second master device corresponding to the slave node module, and obtain the target data broadcast by the initial node module according to the second data read request, Sending the target data to the second master device; wherein the identification information carried in the second data read request is consistent with the identification information carried in the first data read request;
  • the slave device is configured to receive the first data read request sent by the initial node module, and send target data to the initial node module according to the first data read request.
  • the second aspect of the embodiments of the present application provides a data transmission method, which is applied to the initial node module in the broadcast module, and the data transmission method includes:
  • the third aspect of the embodiments of the present application provides another data transmission method, which is applied to the slave node module in the broadcast module, and the data transmission method includes:
  • the fourth aspect of the embodiments of the present application provides a broadcast module. If the broadcast module is specifically an initial node module, it includes:
  • a first receiving unit configured to receive a first data read request sent by a first master device corresponding to the initial node module, where the first data read request carries identification information
  • a data read request sending unit configured to send the first data read request to a corresponding slave device according to the identification information, and receive target data returned by the slave device;
  • a first sending unit configured to send the target data to the first master device and broadcast the target data to the slave node module corresponding to the initial node module;
  • the broadcast module is specifically a slave node module, including:
  • a second receiving unit configured to receive a second data read request sent by a second master device corresponding to the slave node module
  • the target data obtaining unit is configured to obtain, according to the second data reading request, the target data broadcast by the initial node module corresponding to the slave node module, wherein the initial node module contains the first data read carrying the identification information ask;
  • the second sending unit is configured to send the target data to the second master device.
  • the fifth aspect of the embodiments of the present application provides a broadcast module, including:
  • the read command cache submodule is used to access an external main device through the read command channel to obtain a data read request sent by the main device, and the data read request carries identification information;
  • Instruction gate module the input end of the instruction gate module is connected with the output end of the read command buffer submodule, and is used to receive the data read request transmitted by the read command buffer submodule;
  • the output terminal is connected to an external slave device, and is used for if the broadcast module is configured as an initial node module, the output terminal of the instruction gate module is in an open state, and the output terminal of the instruction gate module is switched on according to the identification information carried in the data read request.
  • the external slave device determines the slave device corresponding to the data read request, and sends the data read request to the corresponding slave device; and is also used for if the broadcast module is configured as a slave node module , The output terminal of the instruction gate module is in the closed state, and the outgoing of the data read request is prohibited;
  • a first selector the first input terminal of the first selector is connected to the external slave device, and is used to carry data according to the data read request if the broadcast module is configured as the initial node module
  • the identification information obtains target data from the corresponding slave device; if the broadcast module is configured as a slave node module, the second input terminal of the first selector is configured to be the same as the parent of the broadcast module
  • the node module is connected to obtain target data from the parent node module; wherein the parent node module is another broadcast module located at the upper level node of the broadcast module in the broadcast topology;
  • the read data cache sub-module the input end of the read data cache sub-module is connected to the output end of the first selector, and is used to cache the target data; if the broadcast module has a corresponding child node module, then The output end of the read data cache sub-module is configured to be connected to the sub-node module for sending the target data to the sub-node module; wherein the sub-node module is located in the broadcast topology in the broadcast topology. Other broadcast modules of the next-level node of the module;
  • the second selector, the first input terminal and the second input terminal of the second selector are respectively connected to the read data buffer submodule and the external slave device, and are used to obtain the target data;
  • the output terminal of the second selector is connected to the master device through a read data channel, and is used to transmit the target data to the master device.
  • the sixth aspect of the embodiments of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and running on the processor.
  • the processor executes the computer program
  • the terminal device is caused to implement the steps of the data transmission method as described in the second aspect or the third aspect.
  • the seventh aspect of the embodiments of the present application provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the terminal device realizes the second aspect or The steps of the data transmission method described in the third aspect.
  • the eighth aspect of the embodiments of the present application provides a computer program product, which when the computer program product runs on a terminal device, causes the terminal device to execute the steps of the data transmission method described in the second aspect or the third aspect.
  • the embodiment of this application has the following beneficial effects:
  • the broadcast module is used as the medium for data transmission between the slave device and at least two master devices, and the initial node module and the slave The device interacts, sends a data read request to the slave device and obtains the target data returned from the device; then, through the target data transmission between the initial node module and the slave node module in the broadcast module, the target data is finally transmitted to each The main device.
  • multiple master devices can obtain the target data of the same slave device when the target data is read once, so the data between the master device and the slave device can be improved. Transmission efficiency; and, because in the data transmission process, the slave device only needs to interact with the initial node module in the broadcast module, and does not need to process multiple data read requests repeatedly, thereby reducing system power consumption.
  • FIG. 1 is a structural example diagram of the first system-on-chip provided by an embodiment of the present application
  • FIG. 2 is a structural example diagram of a second system-on-chip provided by an embodiment of the present application
  • FIG. 3 is a structural example diagram of a third system-on-chip provided by an embodiment of the present application.
  • FIG. 4 is a structural example diagram of a fourth system-on-chip provided by an embodiment of the present application.
  • FIG. 5 is a schematic structural diagram of another type of system-on-chip provided by an embodiment of the present application.
  • FIG. 6 is a schematic diagram of the implementation process of the first data transmission method provided by an embodiment of the present application.
  • FIG. 7 is a schematic diagram of the implementation process of the second data transmission method provided by an embodiment of the present application.
  • Fig. 8 is a structural example diagram of an initial node module provided by an embodiment of the present application.
  • FIG. 9 is a structural example diagram of a slave node module provided by an embodiment of the present application.
  • FIG. 10 is an example diagram of the internal structure of a broadcast module provided by an embodiment of the present application.
  • FIG. 11 is a schematic diagram of a terminal device provided by an embodiment of the present application.
  • FIG. 1 shows a schematic structural diagram of the first system on chip provided by an embodiment of the present application, and the details are as follows:
  • the system-on-chip includes at least two master devices 1 (Master), at least one slave device 2 (Slave), and a bus 3 (BUS); the bus 3 includes at least two broadcast modules 31 (Broadcast module, BRC), and each Each of the broadcast modules 31 is connected to one of the main equipment 1; wherein, the broadcast module configured as the initial node in the broadcast topology is the initial node module, and the initial node module is excluded from the broadcast topology.
  • the other broadcast modules are slave node modules.
  • the bus is a common communication backbone for the various functional components in the system-on-chip to transmit information;
  • the master device is a device that has the control right of the bus during a bus transmission, and the slave device is a device that is accessed by the master during a bus transmission.
  • the main equipment provides the corresponding data equipment.
  • the master device may be a processor, and the slave device may be a memory.
  • the system on chip may include M master devices 1, N slave devices 2 and a bus 3, where the value of M is greater than or equal to two, and the value of N is greater than or equal to one.
  • the values of M and N may be the same or different; preferably, the value of M is greater than the value of N, that is, in the system-on-chip, multiple master devices can share the data of the same slave device.
  • the bus 3 is a common communication trunk for transmission of data between the master device and the slave device.
  • the bus 3 includes two or more broadcast modules 31, and each broadcast module 31 is connected to a master device. 1. That is, each broadcast module 31 is connected to a master device 1 that needs to obtain data through broadcast operations.
  • the system-on-chip may also include a master device 1 that does not need to be connected to the broadcast module 31. In this case, the master device 1 The device does not need to share the data of a slave device with other master devices. The master device can independently obtain data directly from the slave device according to its own data read request.
  • connection relationship between the broadcasting modules is determined according to the configured broadcast topology.
  • the broadcast module configured as the initial node in the broadcast topology is the initial node module, such as BRCO as shown in FIG. 1;
  • the broadcast modules other than the initial node module in the structure are slave node modules, such as BRC1, BRC2, and BRC3 as shown in FIG. 1.
  • the initial node module can directly transmit commands and data with the slave device according to the data read request; the slave node module cannot directly interact with the slave device.
  • each group of broadcast topology structures is respectively configured with an initial node module and at least one slave node module, and each initial node module and slave node module is based on the broadcast obtained in advance
  • the configuration information of the topology structure can learn its corresponding subordinate node modules and initial node modules.
  • the main device is used to send data read requests and obtain target data through the broadcast module.
  • each main device connected to the broadcasting module respectively sends a data reading request and obtains target data through the corresponding broadcasting module.
  • the master devices Master0 ⁇ Master3 respectively send their respective data read requests to the broadcast modules BRC0 ⁇ BRC3, and respectively obtain target data from the broadcast modules BRC0 ⁇ BRC3.
  • the initial node module is configured to receive a first data read request sent by a first master device corresponding to the initial node module; according to the identification information carried in the first data read request, the first data The read request is sent to the slave device corresponding to the identification information, and the target data returned by the slave device is received, the target data is sent to the first master device, and the target data is broadcast to all The subordinate node module of the initial node module.
  • the initial node module in the embodiment of the application is the broadcast module of the initial node in the broadcast topology of the bus, and the initial node module is connected to at least one corresponding slave node module; the first master device in the embodiment of the application is the same as the initial node module.
  • the main device connected to the node module.
  • the initial node module After receiving the first data read request sent by the first master device, the initial node module determines the slave device corresponding to the identification information from the plurality of slave devices according to the identification information carried in the first data read request, And send the first data read request to the slave device.
  • the identification information may be the address information of the target data requested by the first data read request, and the address information may include the physical address information of the slave device storing the target data and the requested target data in the slave device.
  • the specific index address information in. Determine the corresponding slave device according to the address information. After sending the first data read request to the slave device, receive the target data returned by the slave device, send the target data to the first master device, and broadcast the target data to the slave node corresponding to the initial node module Module, that is, the slave node module that has been configured to connect with the initial node module.
  • BRC0 is the initial node module, which is connected to the corresponding first master device Master0, and is used to receive the first data read request sent by the first master device Master0; after that, According to the identification information carried in the first data read request, it is determined that the current corresponding slave device is the slave device Slave1, the first data read request is sent to the slave device Slave1 correspondingly, and the target data returned from the slave device Slave1 is received. After that, BRCO broadcasts the received target data to the slave node modules BRC1, BRC2, and BRC3.
  • the initial node module detects that the slave node module has cached the second data read request that is consistent with the identification information of the first data read request, it is determined that the slave node module is ready to receive the target data. Then broadcast the target data to the slave node module.
  • the slave node module is configured to receive a second data read request sent by a second master device corresponding to the slave node module, and obtain the target data broadcast by the initial node module according to the second data read request, The target data is sent to the second master device; wherein the identification information carried in the second data read request is consistent with the identification information carried in the first data read request.
  • the subordinate node modules in the embodiments of this application are broadcast modules other than the initial node in the broadcast topology. Each subordinate node module is directly connected to the initial node module or indirectly establishes communication with the initial node module through other subordinate node modules; implementation of this application
  • the second master device in the example is the master device connected to the slave node module.
  • BRC1, BRC2, and BRC3 are all slave node modules, which are respectively connected to the corresponding second master device Master1, Master2, and Master3 for receiving data read sent by the corresponding second master device. Fetch request.
  • the slave node modules BRC1, BRC2, BRC3 also directly or indirectly communicate with the initial node module BRCO, and are used to obtain the target data broadcast by the initial node module according to the received second data read request. Afterwards, each sends the acquired target data to their corresponding second master devices Master1, Master2, and Master3. Specifically, the second data read request acquired by the slave node module is consistent with the identification information of the first data read request acquired by the initial node module, that is, the slave node module and its corresponding initial node module all receive the same type Data read request to obtain the same target data.
  • each slave node module independently searches for the slave device corresponding to the second data reading request. , Independently obtain the target data corresponding to the second data reading request from the corresponding slave device, instead of acquiring the target data through the initial node module.
  • the slave device is configured to receive the first data read request sent by the initial node module, and send the target data to the initial node module according to the first data read request.
  • the slave device in the embodiment of the present application is specifically a slave device determined by the initial node module according to the identification information carried in the current first data read request.
  • the slave device In the broadcast topology, the slave device only interacts with the initial node module in the broadcast topology. Perform data transfer. Specifically, the slave device receives the first data read request sent by the initial node module, queries the corresponding target data stored in the slave device itself according to the first data read request, and sends the target data to the initial node module.
  • the slave device Slave1 only establishes a connection with the initial node module BRCO, receives a data read request from BRCO, and returns the target data corresponding to the data read request to BRCO.
  • the broadcast module is used as the medium for data transmission between the slave device and at least two master devices, and the initial node module in the broadcast module interacts with the slave device, sends a data read request to the slave device and obtains the slave device.
  • the target data returned by the device; after that, the target data is transmitted to each master device through the target data transmission between the initial node module and the slave node module in the broadcast module. Due to the multiple broadcast modules in the bus, multiple master devices can obtain the target data of the same slave device when the target data is read once, so the data between the master device and the slave device can be improved. Transmission efficiency; and, because in the data transmission process, the slave device only needs to interact with the initial node module in the broadcast module, and does not need to process multiple data read requests repeatedly, thereby reducing system power consumption.
  • the broadcast topology is a two-level topology or a two-level topology
  • the initial node module is configured to be directly connected to each of the slave node modules;
  • the initial node module is configured to be connected to at least one designated slave node module, and there is a preset connection relationship between each of the slave node modules .
  • the broadcast topology in the embodiment of the present application may be a two-level or more than two-level topology.
  • the initial node module in the broadcast topology is configured to be directly connected to each subordinate node module, and there is no connection relationship between each subordinate node module.
  • the initial node module BRCO is directly connected to the subordinate node modules BRC1, BRC2, and BRC3.
  • the broadcast of the target data by the initial node module specifically refers to directly sending the target data to each subordinate node module.
  • FIG. 3 is a schematic diagram of a three-level broadcast topology, where the initial node module BRCO is connected to the slave node module BRC1, and the slave node module BRC1 is further connected to the slave node module BRC2. Connect with the slave node module BRC3.
  • FIG. 3 is a schematic diagram of a three-level broadcast topology, where the initial node module BRCO is connected to the slave node module BRC1, and the slave node module BRC1 is further connected to the slave node module BRC2. Connect with the slave node module BRC3.
  • the initial node module BRCO is connected with the slave node module BRC1
  • the slave node module BRC1 is further connected with the slave node module BRC2
  • the slave node module BRC2 is further connected with the slave node module BRC3, thereby forming a gradual connection relationship.
  • the broadcast of the target data by the initial node module is specifically to send the target data to the subordinate node module directly connected to the initial node module, and then this The subordinate node module transmits the target data to other subordinate node modules that are not connected with the initial node module step by step.
  • the three types of broadcast topology structures in FIG. 2, FIG. 3, and FIG. 4 are only examples and not limitation, and the broadcast topology structures may also be structures not shown in other figures.
  • the hierarchical relationship of the broadcast topology can be specifically set according to actual needs.
  • the broadcast topology is configured as a structure with fewer levels, such as a two-level node structure, to improve the real-time performance of data transmission ;
  • the broadcast topology is configured as a higher-level structure to improve the timing of data transmission.
  • system on chip further includes:
  • the main control module is used to obtain user configuration instructions, generate and transmit the configuration information of the broadcast topology to each of the broadcast modules, where the configuration information of the broadcast topology includes at least the module type of each broadcast module and each of the broadcast modules.
  • the module types include the initial node module and the subordinate node module.
  • the parent node module of the broadcasting module is in the broadcasting topology.
  • Other broadcast modules located at the upper-level node of the broadcast module, and child node modules of the broadcast module are other broadcast modules located at the lower-level node of the broadcast module in the broadcast topology.
  • the system-on-chip shown in FIG. 1 may also include a main control module 4, which can obtain user configuration instructions, generate configuration information of the current broadcast topology, and configure the configuration of the broadcast topology
  • the information is transmitted to each broadcast module of the system-on-chip, so that each broadcast module clarifies its own module type and connection relationship in the corresponding broadcast topology.
  • the configuration information of the broadcast topology includes the module type, the information of the parent node module of the broadcast module, and the information of the child node module.
  • the module type includes an initial node module and a slave node module. Through this module type, each broadcast module can clarify whether it can directly communicate with the corresponding slave device in the broadcast topology.
  • the parent node module of the broadcast module is another broadcast module located at the upper level node of the broadcast module in the broadcast topology
  • the child node module of the broadcast module is located next to the broadcast module in the broadcast topology.
  • the other broadcast modules of the level node through the information of the parent node module and the information of the child node module, can make the connection relationship between each broadcast module and other broadcast modules for subsequent data broadcast transmission between each other.
  • the broadcast topology shown in Fig. 3 is a structure obtained by configuring the broadcast modules BRC0 ⁇ BRC3 as follows according to user configuration instructions: configure the BRC0 module type as the initial node module, and let it establish a connection with the slave device Slave1 , And configure the information of the child node module of BRC0 as BRC1; configure the module type of BRC1 as a slave node module, and configure the information of its parent node module as BRC0, and configure the information of its child node module as BRC2; BRC2 and The module types of BRC3 are configured as slave node modules, and the information of its parent node module is configured as BRC1, and the two slave node modules are used as the end node modules, so the two broadcast modules are no longer configured with child node module information .
  • the configuration information of the broadcast topology can be flexibly generated according to the user configuration instructions to flexibly construct the broadcast topology of the system-on-chip, the flexibility and scalability of data transmission of the system-on-chip can be improved;
  • the configuration information of the broadcast topology is transmitted to each broadcast module, so that each broadcast module clarifies its own module type and connection relationship, so as to accurately realize the broadcast transmission of the target data afterwards.
  • FIG. 5 a schematic structural diagram of another type of system on chip without a broadcast module is given as a comparative example of the system on chip of the embodiment of the present application.
  • the master device Master0 ⁇ Master3 all need to obtain the same target data from the slave device Slave1, they can only interact with Slave1 in turn.
  • Slave1 needs to receive four data read requests and return the target data four times. It can meet the requirements of master devices Master0 ⁇ Master3.
  • the slave device Slave1 only needs to interact with the initial node module, and only needs to receive a data read request once, return the target data once, and then use the broadcast module to complete each
  • the data transmission of the master device can reduce the number of slave device interactions.
  • One interaction of the slave device allows multiple master devices to obtain the target data, which greatly improves the data transmission efficiency and reduces System power consumption.
  • the system on chip in the embodiment of the present application may be artificial intelligence (Artificial Intelligence, AI) chips.
  • AI artificial Intelligence
  • the master device is specifically multiple neural network processors
  • the slave device is specifically a double-rate synchronous dynamic random access memory (Double Data Rate) that stores shared parameters. SDRAM, DDR). Since multiple neural network processors may need to share parameters when running neural network algorithms, DDR is used as a slave device to store and share parameters.
  • DDR is used as a slave device to store and share parameters.
  • multiple neural network processors can be realized Efficiently obtain the shared parameters of the slave device DDR, and can reduce system power consumption and save DDR bandwidth, thereby improving the computing efficiency of the AI chip.
  • FIG. 6 shows a schematic flowchart of a data transmission method provided by an embodiment of the present application.
  • the execution subject of the data transmission method is the initial node module as described in Embodiment 1, and the details are as follows:
  • a first data read request sent by a first master device corresponding to the initial node module is received, where the first data read request carries identification information.
  • the initial node module is connected to the corresponding first master device, and receives a first data read request sent by the first master device, where the first data read request specifically carries identification information.
  • the identification information may include address information of the target data requested by the first data read request, and the address information may include physical address information of the slave device storing the target data and the requested target data in the slave device. Specific index address information in the device.
  • the identification information may also be the device number information of the designated slave device.
  • the first data read request is sent to the corresponding slave device, and the target data returned by the slave device is received.
  • the slave device After receiving the data read request sent by the first master device, determine the corresponding slave device according to the identification information included in the data read request, and send the data read request to the slave device. After that, the target data returned by the slave device is received.
  • the target data is sent to the first master device and the target data is broadcast to the slave node module corresponding to the initial node module.
  • the initial node module After receiving the target data, the initial node module sends the target data to the first master device corresponding to the initial node module, and broadcasts the target data to the slave node module corresponding to the initial node module.
  • the initial node module may determine the corresponding slave node module according to the configuration information of the broadcast topology structure stored by itself.
  • broadcasting the target data to the slave node module corresponding to the initial node module includes: sending the target data to the slave node module directly connected to the initial node module according to the configuration information of the broadcast topology structure stored by itself, that is, to the slave node module directly connected to the initial node module.
  • the child node module corresponding to the initial node module is not sending the target data to the slave node module directly connected to the initial node module according to the configuration information of the broadcast topology structure stored by itself, that is, to the slave node module directly connected to the initial node module.
  • the child node module is a broadcast module located at the next level node of the initial node module in the broadcast topology structure. After that, the child node module determines whether the target data sent by the initial node module needs to be further transferred to more subordinate node modules that are not directly connected to the initial node module according to the configuration information of the broadcast topology structure.
  • the initial node module first obtains the slave node module corresponding to the initial node module, and if it detects that the corresponding slave node module is in a state ready to receive target data, then broadcasts the target data to the corresponding slave node module.
  • the method further includes:
  • step S603 includes:
  • the target data is sent to the first master device and the target data is broadcast to The corresponding slave node module of the initial node module.
  • the initial node module after receiving the first data read request sent by the first master device, the initial node module caches the first data read request. After the initial node module receives the target data, it is detected whether the slave node module corresponding to the initial node module has a second data read request consistent with the identification information of the first data read request cached by the initial node module. When it is detected that the slave node module corresponding to the initial node module already has a second data read request carrying the same identification information, it means that the master device corresponding to the slave node module also needs to obtain the target data, and the slave node module has also received it. The same data read request is ready to receive target data.
  • the target data is sent to the first master device and the target data is broadcast to the slave node module corresponding to the initial node module.
  • the initial node module directly broadcasts the target data to the slave node modules other than the abnormal node module.
  • the abnormal node module is specifically a slave node module that cannot be detected as having the second data reading request within a preset detection time period. Since when there is an abnormal node module, the initial node module can also broadcast the target data to the slave node modules other than the abnormal node module in time without waiting for the data reading status of the abnormal node module, so the data broadcasting performance can be improved. Timeliness and effectiveness.
  • the first data read request is cached, so that after receiving the target data, the first data read request is compared with the second data read request existing in the slave node module as the target.
  • the trigger condition of the data broadcast enables the initial node module to broadcast the target data at an appropriate time, thereby improving the accuracy and effectiveness of data transmission.
  • the method further includes:
  • the configuration information of the broadcast topology structure configured by the main control module is first obtained, and the node type of itself is determined as the initial node module according to the configuration information of the broadcast topology structure, and the initial node module and the initial node module are determined
  • the connection relationship of other broadcast modules specifically, the information of the child node module connected to the initial node module is determined by the configuration information of the broadcast topology structure, and the child node module is located under the initial node module in the broadcast topology structure.
  • the broadcast module specifically determines its own module type as the initial node module type by obtaining the configuration information of the broadcast topology structure and determines the subordinate node modules that have a connection relationship with the initial node module in the broadcast topology structure, so as to be accurate later. Complete the data broadcast operation.
  • the initial node module performs the operations of sending the first data read request to the slave device and receiving the target data, and then the target data is sent to the second master device corresponding to the initial node module and broadcasted to the The slave node module corresponding to the initial node module.
  • multiple master devices can finally obtain the target data through data broadcasting, thus improving the efficiency of data transmission between the master device and the slave device;
  • the slave device only needs to interact with the initial node module and does not need to process multiple data read requests repeatedly, thus reducing system power consumption.
  • FIG. 7 shows a schematic flowchart of another data transmission method provided by an embodiment of the present application.
  • the execution subject of the data transmission method is the slave node module described in Embodiment 1, and the details are as follows:
  • a second data read request sent by a second master device corresponding to the slave node module is received, where the second data request carries identification information.
  • the slave node module is connected to the corresponding second master device, and receives a data read request sent by the second master device.
  • the second data request carries identification information.
  • the identification information may include the second data read request.
  • Address information of the requested target data the address information may include physical address information of the slave device storing the target data and specific index address information of the requested target data in the slave device.
  • the identification information may also be the device number information of the designated slave device.
  • the target data broadcast by the initial node module corresponding to the slave node module is acquired according to the second data read request, where the initial node module has a first data read request carrying the identification information.
  • the target data broadcast by the corresponding initial node module of the slave node module is obtained, where the target data is data corresponding to the second data read request.
  • the slave node module determines the corresponding initial node module according to its pre-stored configuration information of the broadcast topology structure, and the initial node module has a first data read request consistent with the identification information carried in the second data read request.
  • the subordinate node module directly receives the target data sent by the initial node module; if the subordinate node module is not directly connected with the initial node module, but through other intermediate nodes If the slave node module communicates indirectly with the initial node module, the slave node module receives the target data forwarded by its own parent node module (that is, the broadcast module of the upper-level node of the slave node module in the broadcast topology). Indirectly obtain the target data broadcast by the initial node module.
  • the target data is sent to the second master device.
  • the slave node module After receiving the target data, the slave node module at least sends the received target data to the second master device, so that the second master device obtains the data corresponding to the data read request.
  • the slave node module has a corresponding child node module.
  • the step S703 includes:
  • the subordinate node module is specifically a broadcast module that is an intermediate node in a two-level or more broadcast topology, that is, the subordinate node module has a corresponding sub-node module, and the sub-node module is specifically in the broadcast topology.
  • Other broadcast modules located at the next node of the subordinate node module are specifically a broadcast module that is an intermediate node in a two-level or more broadcast topology, that is, the subordinate node module has a corresponding sub-node module, and the sub-node module is specifically in the broadcast topology.
  • Other broadcast modules located at the next node of the subordinate node module.
  • the slave node module in the embodiment of the present application may be the slave node module BRC1 as shown in FIG. 3, which has corresponding child node modules BRC2 and BRC3.
  • the slave node module in the embodiment of the present application may be the slave node module BRC1 as shown in FIG. 4, which has a corresponding sub-node module BRC2; or the slave node module in the embodiment of the present application may be as shown in FIG. 4
  • the slave node module BRC2 has a corresponding child node module BRC3.
  • the slave node module in addition to sending the target data to the second master device corresponding to the slave node module, the slave node module further broadcasts the target data to the child node modules of the slave node module, so that more master devices can pass the broadcast module. Obtain target data.
  • the method further includes:
  • the sending the target data to the second master device and broadcasting the target data to the child node module of the slave node module includes:
  • the slave node module after receiving the second data read request sent by the second master device, the slave node module caches the second data read request. After the slave node module receives the target data, it is detected whether the slave node module of the slave node module has a second data read request consistent with the second data read request cached by the slave node module. When it is detected that the child node module of the slave node module has the same data read request, it means that the master device corresponding to the child node module also needs to obtain the target data, and the child node module also receives the same data read request. Ready to receive target data. At this time, the target data is sent to the second master device and the target data is broadcast to the child node modules of the slave node module.
  • the second data read request is cached, so that after receiving the target data, the second data read request is compared with the second data read request existing in the sub-node module as the target.
  • the trigger condition of the data broadcast enables the slave node module to further broadcast the target data at an appropriate time, thus improving the accuracy and effectiveness of data transmission.
  • the method further includes:
  • the topology information configured by the master control module is obtained, the node type of itself is determined as the slave node module according to the topology information, and the other nodes of the upper level node connected to the slave node module are determined.
  • the information of the broadcast module namely the parent node module
  • the information of other broadcast modules namely the child node modules
  • the unique identification number of the parent node module is used as the information of the parent node module
  • the unique identification number of the child node module is used as the information of the child node module.
  • the information of the sub-node module can be empty, or characters such as "null", "0", "N/A", etc. can be used as the information of the sub-node module .
  • the position of the subordinate node module in the broadcast topology can be accurately determined, so that the subordinate node module can accurately transmit data.
  • the slave node module since neither the slave node module nor the second master device corresponding to the slave node module need to interact with the slave device, only the target data broadcast by the initial node module corresponding to the slave node module can be obtained. 2.
  • the master device obtains the target data. Since the second master device can obtain the target data through the slave node module without waiting for the interaction with the slave device one by one, the data transmission efficiency between the master device and the slave device can be improved; and, because the slave device There is no need to repeatedly process the data read request of the second master device, so the system power consumption can be reduced.
  • FIG. 8 shows a schematic structural diagram of the first broadcast module provided by an embodiment of the present application.
  • the broadcast module is specifically an initial node module.
  • Only the parts related to the embodiment of the present application are shown:
  • the initial node module includes: a first receiving unit 81, a data read request sending unit 82, and a first sending unit 83. in:
  • the first receiving unit 81 is configured to receive a first data read request sent by a first master device corresponding to the initial node module, where the first data read request carries identification information.
  • the data read request sending unit 82 is configured to send the first data read request to the corresponding slave device according to the identification information, and receive the target data returned by the slave device.
  • the first sending unit 83 is configured to send the target data to the first master device and broadcast the target data to the slave node module corresponding to the initial node module.
  • the initial node module further includes:
  • a cache unit configured to cache the first data read request
  • the first sending unit 83 is specifically configured to send the target data to when it is detected that the slave node module corresponding to the initial node module already has a second data read request carrying the identification information
  • the first master device broadcasts the target data to the slave node module corresponding to the initial node module.
  • the initial node module further includes:
  • the first obtaining module is configured to obtain configuration information of the broadcast topology, and determine its own module type and the connection relationship with other broadcast modules according to the configuration information of the broadcast topology.
  • FIG. 9 shows a schematic structural diagram of a second type of broadcast module provided by an embodiment of the present application.
  • the broadcast module is specifically a slave node module.
  • the parts related to the embodiment of the present application are shown:
  • the slave node module includes: a second receiving unit 91, a target data acquiring unit 92, and a second sending unit 93. in:
  • the second receiving unit 91 is configured to receive a second data read request sent by the second master device corresponding to the slave node module.
  • the target data acquiring unit 92 is configured to acquire the target data broadcast by the initial node module corresponding to the slave node module according to the second data read request, wherein the initial node module has a first data read carrying the identification information. Fetch request.
  • the second sending unit 93 is configured to send the target data to the second master device.
  • the slave node module has child node modules.
  • the second sending unit 93 is specifically configured to send the target data to the second master device and broadcast the target data to the The child node module of the subordinate node module.
  • the slave node module further includes:
  • the second acquiring module is used to acquire the configuration information of the broadcast topology, and determine its own module type and the connection relationship with other broadcast modules according to the configuration information of the broadcast topology.
  • the physical structures of the two types of broadcast modules, the initial node module and the subordinate node module, in the embodiments of the present application may be the same or different.
  • the initial node module and the subordinate node module are broadcast modules with different functions obtained by performing different line gating rules on the broadcasting modules of the same physical structure, or by installing different computer programs.
  • FIG. 10 shows a schematic diagram of the physical structure of a broadcast module provided by an embodiment of the present application.
  • the broadcast module includes a read command buffer sub-module ARBUF, a read data buffer sub-module RBUF, an instruction gate module AR CMD Gating, a first selector MUX-0 and a second selector MUX-1, in which:
  • the read command buffer submodule ARBUF is used to access an external main device through the read command channel AR Channel to obtain a data read request sent by the main device, and the data read request carries identification information.
  • the read command buffer submodule ARBUF accesses an external main device through the read command channel AR Channel to obtain a data read request sent by the main device, and the data read request carries identification information.
  • the identification information may include the address information of the target data requested by the data read request, and the address information may include the physical address information of the slave device storing the target data and the specific index address of the requested target data in the slave device. information.
  • the identification information may also be the device number information of the designated slave device.
  • Command gate module AR CMD Gating said command gate module AR The input end of CMD Gating is connected with the output end of the read command buffer submodule ARBUF, and is used to receive the data read request transmitted by the read command buffer submodule ARBUF; the instruction gate module AR The output terminal of CMD Gating is connected with an external slave device, and is used to instruct the gate module AR if the broadcast module is configured as an initial node module
  • the output terminal of CMD Gating is in the on state, and the slave device corresponding to the data read request is determined from the external slave device according to the identification information carried in the data read request, and the data read request Sent to the corresponding slave device; and also used for instructing the gate module AR if the broadcast module is configured as a slave node module
  • the output terminal of CMD Gating is in the closed state, and the outgoing of the data read request is prohibited.
  • the instruction gate module AR CMD Gating of the embodiment of the present application is specifically used to control the outgoing of data read requests, specifically, through the instruction gate module AR The state configuration (including open state and closed state) of the output terminal of CMD Gating controls the outgoing of data read requests.
  • the instruction gate module AR The input terminal of CMD Gating is connected with the output terminal of the read command buffer sub-module ARBUF, and receives the data read request transmitted from the read command buffer sub-module ABRUF.
  • Command gate module AR The output terminal of CMD Gating is connected with external slave devices. Specifically, physical connections can be established in advance with all slave devices outside the broadcast module on the chip system.
  • the output terminal of CMD Gating is configured to be in the open state.
  • the instruction gate module AR CMD Gating receives a data read request, it will send data from all external slaves that have pre-established physical connections according to the identification information carried in the data read request.
  • the device determines the slave device corresponding to the current data read request, and sends the data read request to the corresponding slave device. Specifically, if the broadcast module is configured as a slave node module, the instruction gate module AR The output terminal of CMD Gating is closed, and the read command cache sub-module ABRUF cannot send the data read request.
  • the first selector MUX-0 the first input terminal of the first selector MUX-0 is connected to the external slave device, and is configured to follow the slave device if the broadcast module is configured as the initial node module The corresponding slave device obtains target data; if the broadcast module is configured as a slave node module, the second input terminal of the first selector is configured to be connected to the parent node module of the broadcast module for slave The parent node module obtains target data; wherein, the parent node module is another broadcast module located at the upper-level node of the broadcast module in the broadcast topology.
  • the first input terminal of the first selector MUX-0 establishes a physical connection with all external slave devices in advance.
  • the broadcast module is configured as the initial node module, it corresponds to the identification information carried in the data read request according to the aforementioned determination
  • the slave device obtains the target data from the corresponding slave device.
  • the second input terminal of the first selector MUX-0 communicates with the parent node module of the slave node module (that is, located in the broadcasting module in the broadcasting topology) through the broadcasting line BRC_BUS.
  • the parent node module may be the initial node module, or may be other subordinate node modules between the current subordinate node module and the initial node module.
  • the read data buffer submodule RBUF the input end of the read data buffer submodule RBUF is connected to the output end of the first selector MUX-0, and is used to buffer the target data; if the broadcast module has a corresponding submodule Node module, the output terminal of the read data buffer sub-module RBUF is configured to be connected to the sub-node module and used to send the target data to the sub-node module.
  • the input end of the read data buffer sub-module RBUF is connected to the output end of the first selector MUX-0, and is used to buffer the acquired target data. If the broadcast module itself has a corresponding child node module (that is, other broadcast modules located at the next node of the broadcast module in the broadcast topology), one output end of the read data buffer sub-module RBUF is configured to pass through the broadcast line BRC_BUS is connected to the sub-node module and used to send the target data to the sub-node module.
  • BRC_BUS is connected to the sub-node module and used to send the target data to the sub-node module.
  • the second selector MUX-1, the first input terminal and the second input terminal of the second selector MUX-1 are respectively connected to the read data buffer sub-module RBUF and the external slave device for obtaining all The target data; the output terminal of the second selector MUX-1 is connected to the master device through the read data channel R Channel, and is used to transmit the target data to the master device.
  • the first input terminal of the second selector MUX-1 is connected to the read data buffer sub-module RBUF, and the second input terminal of the second selector MUX-1 is connected to an external slave device to obtain target data.
  • the second selector selects to obtain the target data from the read data buffer submodule RBUF through the first input; when the broadcast module does not need to perform the target data broadcast operation, the second selector Select to directly obtain the target data of the slave device corresponding to the identification information of the current data reading request through the second input terminal.
  • the output terminal of the second selector MUX-1 is connected to the corresponding master device through the read data channel R Channel to transmit target data to the master device.
  • the read command cache sub-module ABRUF is further configured to connect to the parent node module through the broadcast line BRC_BUS if the broadcast module is configured as a slave node module, so as to send the data read request to the parent node after it has obtained the data read request.
  • the module informs itself that it is ready to receive target data.
  • the read command buffer sub-module ABRUF is also used to connect to the sub-node module through the broadcast line BRC_BUS if there is a sub-node module in the broadcast module to receive the notification of the ready state of the sub-node module; when the next-level module is received When the node's broadcast module is notified of the ready status, the broadcast line BRC_BUS is then controlled to transmit the target data to the child node module.
  • the above-mentioned structure of the broadcast module is used to select the line, and the module type and the corresponding connection relationship of the broadcast module can be flexibly configured, so as to be used as the medium between the master device and the slave device to flexibly realize the corresponding Data transfer function, thereby improving the data transfer efficiency between the master device and the slave device and reducing the system power consumption of the system-on-chip.
  • FIG. 11 is a schematic diagram of a terminal device provided by an embodiment of the present application.
  • the terminal device 11 of this embodiment includes: a processor 110, a memory 111, and a computer program 112 stored in the memory 111 and running on the processor 110, such as a data transfer program.
  • the processor 110 executes the computer program 112
  • the steps in the foregoing data transmission method embodiments are implemented, such as steps S601 to S603 shown in FIG. 6 or steps S701 to S703 shown in FIG. 7.
  • the processor 110 executes the computer program 112
  • the functions of the modules/units in the foregoing device embodiments are implemented, such as the functions of the units 81 to 83 shown in FIG. 8 or the functions of the units 91 to 93 shown in FIG.
  • FIG. 11 is a schematic diagram of a terminal device provided by an embodiment of the present application.
  • the terminal device 11 of this embodiment includes: a processor 110, a memory 111, and a computer program 112 stored in the memory 111
  • the computer program 112 may be divided into one or more modules/units, and the one or more modules/units are stored in the memory 111 and executed by the processor 110 to complete This application.
  • the one or more modules/units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program 112 in the terminal device 11.
  • the computer program 112 can be divided into a first receiving unit, a data reading request sending unit, and a first sending unit, and the specific functions of each unit are as follows:
  • the first receiving unit is configured to receive a first data read request sent by a first master device corresponding to the initial node module, where the first data read request carries identification information.
  • the data read request sending unit is configured to send the first data read request to the corresponding slave device according to the identification information, and receive the target data returned by the slave device.
  • the first sending unit is configured to send the target data to the first master device and broadcast the target data to the slave node module corresponding to the initial node module.
  • the computer program 112 may be divided into a second receiving unit, a target data acquiring unit, and a second sending unit, and the specific functions of each unit are as follows:
  • the second receiving unit is configured to receive a second data read request sent by the second master device corresponding to the slave node module.
  • the target data obtaining unit is configured to obtain, according to the second data reading request, the target data broadcast by the initial node module corresponding to the slave node module, wherein the initial node module contains the first data read carrying the identification information ask.
  • the second sending unit is configured to send the target data to the second master device.
  • the terminal device 11 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server.
  • the terminal device may include, but is not limited to, a processor 110 and a memory 111.
  • FIG. 11 is only an example of the terminal device 11, and does not constitute a limitation on the terminal device 11. It may include more or less components than shown in the figure, or a combination of certain components, or different components.
  • the terminal device may also include input and output devices, network access devices, buses, and so on.
  • the so-called processor 110 may be a central processing unit (Central Processing Unit, CPU), other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like.
  • the memory 111 may be an internal storage unit of the terminal device 11, such as a hard disk or a memory of the terminal device 11.
  • the memory 111 may also be an external storage device of the terminal device 11, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), and a secure digital (Secure Digital, SD) equipped on the terminal device 11. Flash card Card) etc.
  • the memory 111 may also include both an internal storage unit of the terminal device 11 and an external storage device.
  • the memory 111 is used to store the computer program and other programs and data required by the terminal device.
  • the memory 111 can also be used to temporarily store data that has been output or will be output.

Abstract

La présente invention s'applique au domaine technique de l'électronique et concerne un système sur puce, un procédé de transmission de données et des modules de diffusion. Le système sur puce comprend au moins deux dispositifs maîtres, au moins un dispositif esclave et un bus ; le bus comprend au moins deux modules de diffusion, et chaque module de diffusion est connecté de manière correspondante à un dispositif maître ; et les modules de diffusion sont configurés en tant que modules de nœuds initiaux ou modules de nœuds esclaves. Selon des modes de réalisation de la présente invention, les modules de diffusion sont utilisés en tant que supports pour la transmission de données entre le dispositif esclave et lesdits deux dispositifs maîtres dans le système sur puce ; au moyen d'une interaction entre les modules de nœuds initiaux dans les modules de diffusion et le dispositif esclave, une demande de lecture de données est envoyée au dispositif esclave, et des données cibles renvoyées par le dispositif esclave sont obtenues ; ensuite, les données cibles sont transmises à chaque dispositif maître au moyen de la transmission de données cibles entre les modules de nœuds initiaux et les modules de nœuds esclaves dans les modules de diffusion. Les modes de réalisation de la présente invention peuvent résoudre les problèmes de l'amélioration de l'efficacité de transmission de données du système sur puce et de réduction de la consommation d'énergie dans l'état de la technique.
PCT/CN2021/088442 2020-06-04 2021-04-20 Système sur puce, procédé de transmission de données et modules de diffusion WO2021244168A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010500891.2A CN111797051B (zh) 2020-06-04 2020-06-04 片上系统、数据传送方法及广播模块
CN202010500891.2 2020-06-04

Publications (1)

Publication Number Publication Date
WO2021244168A1 true WO2021244168A1 (fr) 2021-12-09

Family

ID=72803009

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/088442 WO2021244168A1 (fr) 2020-06-04 2021-04-20 Système sur puce, procédé de transmission de données et modules de diffusion

Country Status (2)

Country Link
CN (1) CN111797051B (fr)
WO (1) WO2021244168A1 (fr)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111797051B (zh) * 2020-06-04 2022-05-17 深圳云天励飞技术股份有限公司 片上系统、数据传送方法及广播模块
CN112822127B (zh) * 2020-12-29 2023-07-04 北京灵汐科技有限公司 片上网络的数据广播方法、装置、芯片及介质
CN112367236B (zh) * 2021-01-12 2021-04-06 南京芯驰半导体科技有限公司 Lin总线的数据调度方法及系统
WO2022246803A1 (fr) * 2021-05-28 2022-12-01 华为技术有限公司 Procédé d'accès et appareil de communication

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107341129A (zh) * 2016-04-29 2017-11-10 上海磁宇信息科技有限公司 细胞阵列计算系统及其测试方法
CN108121679A (zh) * 2017-08-07 2018-06-05 鸿秦（北京）科技有限公司 一种嵌入式SoC系统总线及其协议转换桥接装置
CN108132910A (zh) * 2016-12-01 2018-06-08 三星电子株式会社 系统互连以及具有系统互连的片上系统
CN110674075A (zh) * 2019-09-27 2020-01-10 山东华芯半导体有限公司 一种axi总线广播机制的实现方法和系统
CN111797051A (zh) * 2020-06-04 2020-10-20 深圳云天励飞技术有限公司 片上系统、数据传送方法及广播模块

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH05158891A (ja) * 1991-12-06 1993-06-25 Fujitsu Ltd 同報送信における応答情報収集方式
CN100365602C (zh) * 2004-12-31 2008-01-30 北京中星微电子有限公司 实现多个主动装置对单一总线上从动装置进行存取的设备
US7246188B2 (en) * 2005-02-10 2007-07-17 Qualcomm Incorporated Flow control method to improve bus utilization in a system-on-a-chip integrated circuit
US7617343B2 (en) * 2005-03-02 2009-11-10 Qualcomm Incorporated Scalable bus structure
JP2007241612A (ja) * 2006-03-08 2007-09-20 Matsushita Electric Ind Co Ltd マルチマスタシステム
US7493468B2 (en) * 2006-06-01 2009-02-17 International Business Machines Corporation Method for broadcasting instructions/data to a plurality of processors in a multiprocessor device via aliasing
KR100813256B1 (ko) * 2006-06-23 2008-03-13 삼성전자주식회사 버스 중재 장치 및 방법
CN100524252C (zh) * 2007-09-20 2009-08-05 华为技术有限公司 一种嵌入式系统芯片及数据读写处理方法
KR20120097831A (ko) * 2011-02-25 2012-09-05 삼성전자주식회사 시스템 온 칩의 버스 시스템
US9009446B2 (en) * 2012-08-02 2015-04-14 Oracle International Corporation Using broadcast-based TLB sharing to reduce address-translation latency in a shared-memory system with electrical interconnect

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107341129A (zh) * 2016-04-29 2017-11-10 上海磁宇信息科技有限公司 细胞阵列计算系统及其测试方法
CN108132910A (zh) * 2016-12-01 2018-06-08 三星电子株式会社 系统互连以及具有系统互连的片上系统
CN108121679A (zh) * 2017-08-07 2018-06-05 鸿秦（北京）科技有限公司 一种嵌入式SoC系统总线及其协议转换桥接装置
CN110674075A (zh) * 2019-09-27 2020-01-10 山东华芯半导体有限公司 一种axi总线广播机制的实现方法和系统
CN111797051A (zh) * 2020-06-04 2020-10-20 深圳云天励飞技术有限公司 片上系统、数据传送方法及广播模块

Also Published As

Publication number Publication date
CN111797051A (zh) 2020-10-20
CN111797051B (zh) 2022-05-17

Similar Documents

Publication Publication Date Title
WO2021244168A1 (fr) Système sur puce, procédé de transmission de données et modules de diffusion
US20210112003A1 (en) Network interface for data transport in heterogeneous computing environments
EP2985699B1 (fr) Procédé d'accès à la mémoire et système de mémoire
US8699953B2 (en) Low-latency interface-based networking
WO2020078470A1 (fr) Procédé et dispositif de traitement de données réseau sur puce
US9576026B2 (en) System and method for distributed SQL join processing in shared-nothing relational database clusters using self directed data streams
WO2022099573A1 (fr) Procédé d'extension de puce pour ordinateur de type cerveau de niveau cent millions
CN105874758B (zh) 内存访问方法、交换机及多处理器系统
US20120044942A1 (en) Semiconductor system, relay apparatus, and chip circuit
CN111630487A (zh) 用于神经网络处理的共享存储器的集中式-分布式混合组织
WO2023207035A1 (fr) Procédé et appareil de synchronisation de données, dispositif et support de stockage
CN108270877B (zh) 分布式网络节点数据共享系统
CN107391422A (zh) 多路异步串行通讯数据访问系统及方法
CN113190528B (zh) 一种并行分布式大数据架构构建方法及系统
US20230403232A1 (en) Data Transmission System and Method, and Related Device
CN102591817B (zh) 一种多总线桥控制器及其实现方法
CN116301667A (zh) 一种数据库系统，数据访问方法、装置、设备及存储介质
US10496565B2 (en) Micro-architectural techniques to minimize companion die firmware loading times in a server platform
US20220358078A1 (en) Integrated circuit, data processing device and method
TW202008172A (zh) 儲存系統
WO2021238771A1 (fr) Procédé d'accès à des données en nœud de périphérie, appareil et système
WO2022110681A1 (fr) Procédé de retour et appareil de commande de retour pour informations de réponse de commande, et dispositif électronique
CN112506824A (zh) 一种芯片和数据交互方法
CN112463680A (zh) 数据搬运方法及装置
CN117632820B (zh) 请求处理方法、装置、总线桥、电子设备及可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21817769

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21817769

Country of ref document: EP

Kind code of ref document: A1