CN111767129B - Data flow task processing device and method - Google Patents
Data flow task processing device and method Download PDFInfo
- Publication number
- CN111767129B CN111767129B CN202010613328.6A CN202010613328A CN111767129B CN 111767129 B CN111767129 B CN 111767129B CN 202010613328 A CN202010613328 A CN 202010613328A CN 111767129 B CN111767129 B CN 111767129B
- Authority
- CN
- China
- Prior art keywords
- data
- input
- task
- data flow
- flow task
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Classifications
-
- 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
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- 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
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The application provides a data flow task processing device and a method, wherein the device comprises: the processing module is used for processing data; and the data flow task activation module is used for detecting whether the input data of the data flow task is ready to be finished or not, and controlling the processing module to process the data of the data flow task after the data flow task is ready to be finished. The data flow task processing device monitors the preparation condition of input data needed by a data flow task, calls the data flow task to process when the input data of the data flow task is monitored to be prepared, monitors the hardware and directly triggers the execution of the data flow task, and does not need to read the data flow calculation running state stored in a memory one by one, so that the activation speed of the data flow task is accelerated, and the processing speed of the data flow task is increased.
Description
Technical Field
The invention belongs to the technical field of data processing, and particularly relates to a data flow task processing device and method.
Background
At present, the development direction of processors has been developed from the direction of simply increasing the running speed of the processors to the direction of multi-core processors, and large-scale distributed systems are increasingly common. Conventionally, programming is performed by using a structure of sequentially executing commands, and in this mode, data is often "static" and a continuous access operation is required to the data. This makes programs not particularly well supported by multi-core processors and large distributed systems. And the data flow programming emphasizes that the data is used as driving power, and the connection operation of input and output is clearly defined. And a command mode is not adopted, and related operations are immediately executed when data are ready and input is effective, so that the data flow programming is parallel in nature and can be well operated in a multi-core processor and a large-scale distributed system.
Currently, operations related to the data flow calculation operation, such as the judgment and scheduling of a data flow task, are performed by a program on an operating system by reading a data flow calculation operation state stored in a memory. However, in future large computers, not only the number of processor cores is large, but also computing components in the memory of the IMC (the execution time of each task is very short), the triggering time of scheduling is very critical. The data processing speed of the optical dependence program is judged by reading the memory, and the requirement of data flow calculation cannot be met.
Disclosure of Invention
The invention provides a data stream task processing device and a data stream task processing method, which are used for solving the technical problem that the processing speed is lower when a current program judges and schedules a data stream task by reading the running state of a data stream stored in a memory.
A first aspect of the present invention provides a data stream task processing apparatus, including:
the processing module is used for processing data; and the number of the first and second groups,
and the data stream task activation module is used for detecting whether input data of the data stream task is ready to be finished or not, and controlling the processing module to perform data processing on the data stream task after the data stream task is ready to be finished.
Further, the apparatus further comprises:
and the input data monitoring module is used for monitoring whether data are input or not, and reminding the data stream task activation module of finishing the current data preparation when the data are input.
Further, the input data listening module comprises:
a monitoring address determining submodule for determining a monitoring address according to the data stream task;
the monitoring submodule is used for monitoring whether the address has data input;
and the reminding sub-module is used for reminding the data flow task activation module when data are input into the address.
Further, the data flow task activation module includes:
the input counting submodule is used for counting the data monitored by the data input monitoring module;
and the control submodule controls the processing module to perform data processing on the data stream task when the counting number of the input counting submodule reaches a preset number.
A second aspect of the present application provides a data stream task processing method, which is applied to the data stream task processing device provided by the first aspect, and the method includes:
the data flow task activation module detects whether input data of a data flow task is ready to be completed;
and if the input data of the data stream task is prepared, processing the data stream task by a processing module.
Further, the data flow task activation module detects whether data flow task input data is ready to be completed, including:
when the data stream task input data preparation is completed, the data stream task activation module counts;
and when the numerical value counted by the data flow task activation module reaches a preset numerical value, determining that the data preparation of the data flow task is finished.
As can be seen from the above description, the data stream task processing device provided in the present application includes: the processing module is used for processing data; and the data flow task activation module is used for detecting whether the input data of the data flow task is ready to be finished or not, and controlling the processing module to process the data of the data flow task after the data flow task is ready to be finished. The data flow task processing device monitors the preparation condition of input data required by the data flow task, and calls the data flow task to process when the input data of the data flow task is monitored to be prepared. The data stream computing running state stored in the memory does not need to be read one by one, so that the activation speed of the data stream task is accelerated, and the processing speed of the data stream task is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a schematic structural diagram of a data flow task processing device according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a data flow task processing device according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of an input data monitoring module according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a data flow task activation module according to an embodiment of the present application;
fig. 5 is a schematic flowchart of a data flow task processing method according to an embodiment of the present application;
fig. 6 is a data flow processing directed graph provided in an embodiment of the present application;
fig. 7 is a schematic diagram of a hardware implementation of a part of a data flow task processing device according to an embodiment of the present application.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, a schematic structural diagram of a data stream task processing device provided by the present application is provided, where the device includes the following modules:
a processing module 101, configured to perform data processing; and the number of the first and second groups,
the data flow task activation module 102 is configured to monitor whether input data of the data flow task is ready to be completed, and control the processing module to perform data processing on the data flow task after the data flow task is ready to be completed.
Data flow programming is a high performance parallel programming model that addresses the problem of efficiency utilization of multi-core processors. The data flow programming is obviously different from the traditional programming language, the data flow programming separates the calculation and the communication of data through a data driving mode, potential parallelism in a flow program is fully mined by using the parallel characteristic of software pipelining through task scheduling and distribution and data transmission to a core where a task is located, and load balance among cores is realized. In the data flow paradigm, a static instance of a data flow program will be described as a directed graph in terms of its structure. In the figure, the nodes represent calculation units and the edges represent data transmission paths. And transmitting data between adjacent nodes through edges, calculating the node consumption data, and outputting the generated data to an input-output sequence as the input of the next calculation unit.
In the embodiment of the present application, the processing module 101 of the data flow task processing device is configured to perform data processing, i.e., a data calculation function, in the node of the directed graph. The data flow task activation module 102 is configured to monitor data required for data computation of the current node to determine whether the data required for use is ready to be completed. The fact that the data is ready means that the output nodes of all data required by the operation of the current node are already operated and the operation result is transmitted to the current node through the edge of the directed graph. For example, the current node needs three data to operate, wherein one data is a constant and can be directly transmitted to the current node. One is to square a certain data and the other is to invert a certain data. Then it is only determined that the input data of the dataflow task of the current node is ready to be completed if both operations are calculated and output data is transmitted to the current node. The dataflow task activation module 102 may update its state as each input datum is received and set to reach a set state when the input datum is fully prepared. When the data flow task activation module 102 reaches the set state, the control processing module performs data processing on the data flow task. That is, when the data flow task activation module 102 reaches the set state, the control processing module 101 performs data calculation on the current node.
As can be seen from the above description, the data stream task processing device provided in the present application includes: the processing module is used for processing data; and the data flow task activation module is used for detecting whether the input data of the data flow task is ready to be finished or not, and controlling the processing module to process the data of the data flow task after the data flow task is ready to be finished. The data flow task processing device monitors the preparation condition of input data required by the data flow task, and calls the data flow task to process when the input data of the data flow task is monitored to be prepared. The data stream computing running state stored in the memory does not need to be read one by one, so that the activation speed of the data stream task is accelerated, and the processing speed of the data stream task is improved.
Further, as shown in fig. 3, for another schematic structural diagram of the data stream task processing device provided by the present application, the device further includes:
the input data monitoring module 201 is configured to monitor whether there is data input, and when there is data input, remind the data stream task activation module 102 that the current data preparation is completed.
In the embodiment of the present application, whether all data required for the current node calculation is ready is monitored by the data stream task activation module, and whether each data required for the current node calculation is ready is monitored by the input data monitoring module 201. In the processor, the result of the operation of each node has a storage location, and the input data of each node also has a storage location. And transmitting the data obtained by the operation of the previous node to the input data storage position of the next node for calling when the next node operates. The input data monitoring module 201 may monitor the input data storage location of the current node, and when the input data storage location of the current node is monitored to have data input, a prompt is sent to the data stream task activation module 102. The dataflow task activation module 102 updates its state accordingly. It is understood that the input data monitoring module 201 monitors not only one but all input data storage locations of the current node, and whenever there is data input at one storage location, the input data monitoring module 201 may want the data stream task activation module 102 to send a prompt.
Further, as shown in fig. 3, for a schematic structural diagram of the input data monitoring module provided in the present application, the input data monitoring module includes:
a monitoring address determining submodule 301, configured to determine a monitoring address according to the data stream task;
a monitoring submodule 302, configured to monitor whether data is input into the determined monitoring address;
and the reminding submodule 303 is configured to remind the data stream task activation module when the determined monitoring address has data input.
In the embodiment of the present application, since the data flow task is often multiple or even numerous, and the processing of the data flow task is often continuous. When the input data of a data flow task is ready, the data flow task starts to run. And the input data monitoring module starts to monitor the input data of the next data stream task. Before snooping, the snoop address determining sub-module 301 determines a storage address of input data of the data stream task according to the snooped data stream task, that is, determines a snoop address. After the snoop address is re-determined, the snoop submodule 302 snoops the re-determined snoop address to determine whether data is input. When the redetermined monitoring address is monitored to have data input, the reminding sub-module 303 sends a reminder to the data stream task activating module 102.
Further, as shown in fig. 4, for a schematic structural diagram of a data flow task activation module in the data flow task processing apparatus provided by the present application, the data flow task activation module includes:
an input counting submodule 401, configured to count data monitored by the data input monitoring module;
and the control submodule 402 is configured to control the processing module to process the data stream task when the count number input to the count submodule reaches a preset number.
In this embodiment, the data stream task activation module has an input counting sub-module 401, and when the input data monitoring module monitors that there is data input and sends a prompt to the data stream task activation module, the input counting sub-module counts. The input counting submodule 401 may set a preset value of the input counting submodule according to the number of the input data of the dataflow task, for example, if the number of the input data of the dataflow task is 3, the preset value of the input counting submodule may be set to 3, and when the number counted by the input counting submodule 401 reaches 3, it indicates that all 3 input data of the dataflow task are ready to be completed. At this time, the control sub-module 402 of the data flow task activation module 102 controls the processing module 101 to call the data flow task and process the data flow task. It will be appreciated that the technique of entering the count submodule 401 may also be count-back. Namely, the initial value of the counting submodule is set to 3, and every time the data monitoring module monitors that one data is ready and sends a prompt to the data monitoring module, the input counting submodule subtracts 1 numerical value. It is understood that the data flow task processing device may be disposed on one core of the processor, or may be disposed on each processing core of the multi-core processor, so as to further increase the processing speed of the data flow task.
As shown in fig. 5, a schematic flow chart of a data flow task processing method provided in the present application is shown, where the data flow task processing method can be used in the foregoing data flow task processing apparatus, and the method includes:
Further, the data flow task activation module detects whether data flow task input data is ready to be completed, including:
when the data stream task input data preparation is completed, the data stream task activation module counts;
and when the numerical value counted by the data flow task activation module reaches a preset numerical value, determining that the data flow task input data is ready to be finished.
It can be understood that the content of each step in the data stream task processing method provided in this embodiment is the same as the function of each module in the data stream task processing apparatus in the foregoing embodiment, and is not described herein again.
The following describes the data flow task processing apparatus and method in a specific embodiment:
as shown in fig. 6, for a data flow processing directed graph provided in the embodiment of the present application, as shown in the figure, all nodes in the directed graph may be classified into four types according to their states: a completed node F, a running node R, a node to be run N adjacent to the running node, and an unprepared node U. The input data of the dataflow task C to be executed in the figure are the output data of the dataflow tasks A, B and 3, 4. And the data flow task C is in an N state to be operated, and the operation result of the data flow task A in the input data is not completed. The data stream task activation module 102 and the input data snooping module 201 may be disposed in the processor in the form of a hardware data stream task activation table and an input data snooping table. Namely, the architecture of the processor is modified, and the two components are added outside the components such as cache, TLB and the like outside the core of the original processor and are used for supporting the activation of the data flow task.
As shown in table 1, for the hardware data flow task activation table, each entry records an input data count of an N-type data flow task and its thread number.
Table 1 data flow task activation table
In the table, when the status of the N-type dataflow task is not changed to be in operation, i.e., the N-type dataflow task is not changed to be R-type, the valid column of table 1 is shown as 1, and the input count column of table 1 performs input counting. And each time the input data monitoring table monitors that one data in the input data of the N-type data stream task is ready, the input counting column value is decreased by 1. When the column count is decremented to 0, it is determined that the input data of the N-type dataflow task is ready to be completed, and the state of the N-type dataflow task is changed to R. When the status of the N-type data stream task is changed to R, the active column in table 1 is automatically switched to 0, which indicates that the row count task is invalid. The input data chain in table 1 shows the numbers of the input data snoop table in the following table (i.e., table 2). Each number represents a data to be heard.
Table 2 input data snoop table
It will be appreciated that the number M of the numbers in table 2 is greater than the number N of the numbers in table 1, and that in general 1 entry in table 1 corresponds to at least two entries in table 2. When the valid column in table 1 is changed from 1 to 0, the entries in table 2 are determined from the input data chain in table 1 and the valid columns of these entries are changed from 1 to 0. In the snoop address column of table 2, the address range stored in the data stream data object, i.e. the storage address of the input data of the N-type data stream to be run, is recorded. The data stream task number in table 2 records the number data corresponding to the item in table 1 in table 2.
Thus, when the state of a data flow task is changed from U to N, table 2 obtains the input data storage address of the data flow task, each address corresponds to an entry in table 2, a free entry is allocated in table 2 to input the obtained storage address, and the valid column of each entry in table 2 is set to 1. The above items form a linked list through the 'link pointer', and the 'link pointer' of the last condition data is NULL, which indicates the end of the linked list. And then monitoring input data of the address, sending a prompt to the table 1 when data input is monitored, decrementing the input count of the table 1 by 1, setting the effective column of the table 1 as 1, inputting the item label of the table 2 into the input data chain of the table 1, and writing the thread PID into the thread number. Meanwhile, table 2 fills in the task number of the data stream according to the item number of table 1.
When the input count column in table 1 is decremented to 0, it is determined that the input data of the dataflow task is ready to be completed, the execution of the N-type dataflow task is triggered, and the state thereof is switched to R. Then, all the valid columns in table 1 and table 2 corresponding to the data flow task are set to 0.
According to the above description, the data stream task activation module and the input data snooping module of the data stream task processing device provided by the present application can be represented by two independent internal hardware tables of the processor (analogy)
cache/TLB, with parallel compare capability). The data flow task activation module manages data flow tasks by adopting a data flow task activation table, and the input data monitoring module manages monitored addresses by adopting an input data monitoring table. The entries of the input data snoop table are each uniquely associated with an entry in the data stream task activation table, which is uniquely associated with the unique entry in the input data snoop table corresponding to the first input data of the task. The entries belonging to the same task in the input data monitoring table are linked into a single chain.
Fig. 7 is a schematic diagram of a partial hardware implementation of the data stream task processing apparatus provided in the present application, and as shown in the diagram, a schematic structural diagram of an input data snoop table and a data stream task activation table is shown. The memory write operation sent by the processor is compared in parallel by the input data monitoring table, when a matching item is found, the corresponding 'data stream number' is used as a selection signal of the data stream task activation table, one of the data stream number is selected and the 'input count' value is decreased, when the input count value is decreased to 0, an interrupt is triggered, and the hardware scheduler is triggered to execute the corresponding data stream task.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In view of the above description of the technical solutions provided by the present invention, those skilled in the art will recognize that there may be variations in the technical solutions and the application ranges according to the concepts of the embodiments of the present invention, and in summary, the content of the present specification should not be construed as limiting the present invention.
Claims (1)
1. A method applied to a data stream task processing device, the device comprising:
the processing module is used for processing data; and the number of the first and second groups,
a data flow task activation module, configured to monitor whether input data of a data flow task is ready to be completed, and control the processing module to perform data processing on the data flow task after the data flow task is ready to be completed, where the apparatus further includes:
the input data monitoring module is used for monitoring whether data are input or not, and reminding the data stream task activation module of finishing the current data preparation when the data are input;
the input data listening module comprises:
a monitoring address determining submodule for determining a monitoring address according to the data stream task;
the monitoring submodule is used for monitoring whether the address has data input;
the reminding submodule is used for reminding the data flow task activation module when data are input into the address;
the input counting submodule is used for counting the data monitored by the data input monitoring module;
the control submodule controls the processing module to process the data stream task when the counting number of the input counting submodule reaches a preset number;
the method comprises the following steps:
the data flow task activation module monitors whether input data of the data flow task is ready to be completed or not; when the data stream task input data preparation is completed, the data stream task activation module counts;
when the numerical value counted by the data flow task activation module reaches a preset numerical value, determining that the data flow task input data preparation is finished;
if the input data of the data flow task is prepared, a processing module processes the data flow task;
the data stream task activation module and the input data monitoring module are arranged in a processor in the form of a hardware data stream task activation table and an input data monitoring table;
table 1 is a hardware data flow task activation table, and each entry records an input data count and a thread number of an N-type data flow task;
table 1 data flow task activation table
In the table, when the status of the N-type dataflow task is not changed to be in operation, i.e., the N-type dataflow task is not changed to be R-type, the valid column of table 1 is shown as 1, and the input count column of table 1 performs input counting; when an input data monitoring table monitors that one data in the input data of the N-type data stream task is ready, the input counting column numerical value is decreased by 1; when the column number is decreased to 0, determining that the input data of the N-type data stream task are all ready to be completed, and changing the state of the N-type data stream task to R; when the state of the N-type data stream task is changed into R, the effective column in the table 1 is automatically switched to 0, which indicates that the row counting task is invalid; wherein N represents a node to be operated adjacent to the node in operation, and R represents the node in operation; the input data chain in table 1 records the number of the input data monitoring table in table 2; each number represents a monitored datum;
table 2 input data snoop table
The number M of the numbers in the table 2 is larger than the number of the N in the table 1, and 1 item in the table 1 corresponds to at least two items in the table 2; when the effective column in the table 1 is changed from 1 to 0, determining the items in the table 2 according to the input data chain in the table 1, and changing the effective columns of the items from 1 to 0; in the monitoring address column of table 2, the address range stored in the data stream data object, that is, the storage address of the input data of the N-type data stream to be operated, is recorded; the data stream task number in table 2 records the number data corresponding to the item in table 1 in table 2;
when the state of a data flow task is changed from U to N, the input data storage address of the data flow task is obtained in the table 2, each address corresponds to one item in the table 2, an idle item is allocated in the table 2 to input the obtained storage address, and the effective column of each item in the table 2 is set to be 1; the linked list is formed by the linked pointers, and the linked pointer of the last condition data is NULL to indicate the end of the linked list; then, input data monitoring is carried out on the address, when data input is monitored, a prompt is sent to the table 1, the input count of the table 1 is decreased by 1, the effective column of the table 1 is set to be 1, the item label of the table 2 is input into the input data chain of the table 1, and the thread PID is written into a thread number; meanwhile, the data stream task number is filled in according to the project number in the table 1 in the table 2; wherein U represents an unprepared node;
when the input counting array in the table 1 is decreased to 0, determining that the input data of the data stream task is ready to be completed, triggering the N-type data stream task to be executed, and switching the state of the N-type data stream task to be R; then, all the valid columns in table 1 and table 2 corresponding to the data flow task are set to 0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010613328.6A CN111767129B (en) | 2020-06-30 | 2020-06-30 | Data flow task processing device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010613328.6A CN111767129B (en) | 2020-06-30 | 2020-06-30 | Data flow task processing device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111767129A CN111767129A (en) | 2020-10-13 |
| CN111767129B true CN111767129B (en) | 2021-10-29 |
Family
ID=72723281
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010613328.6A Active CN111767129B (en) | 2020-06-30 | 2020-06-30 | Data flow task processing device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111767129B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106649352A (en) * | 2015-10-30 | 2017-05-10 | 北京国双科技有限公司 | Data processing method and apparatus |
| CN107066318A (en) * | 2017-01-13 | 2017-08-18 | 广西电网有限责任公司电力科学研究院 | A kind of data dispatching method suitable for stream calculation |
| CN107169854A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | A kind of method and device of data processing |
| CN109634766A (en) * | 2019-02-20 | 2019-04-16 | 深圳大学 | Promote method, apparatus, equipment and the storage medium of data flow computer operational efficiency |
| CN109800129A (en) * | 2019-01-17 | 2019-05-24 | 青岛特锐德电气股份有限公司 | A kind of real-time stream calculation monitoring system and method for processing monitoring big data |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8296764B2 (en) * | 2003-08-14 | 2012-10-23 | Nvidia Corporation | Internal synchronization control for adaptive integrated circuitry |
| US7669081B2 (en) * | 2006-09-27 | 2010-02-23 | Raytheon Company | Systems and methods for scheduling, processing, and monitoring tasks |
| CN101739242B (en) * | 2009-11-27 | 2013-07-31 | 深圳中微电科技有限公司 | Stream data processing method and stream processor |
| US8612985B2 (en) * | 2010-11-24 | 2013-12-17 | Siemens Aktiengesellschaft | Apparatuses and methods for processing workitems in taskflows |
| US20130155078A1 (en) * | 2011-12-15 | 2013-06-20 | Ati Technologies Ulc | Configurable graphics control and monitoring |
| CN103853601A (en) * | 2012-11-29 | 2014-06-11 | 杭州勒卡斯广告策划有限公司 | Workflow task creating method and device |
-
2020
- 2020-06-30 CN CN202010613328.6A patent/CN111767129B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106649352A (en) * | 2015-10-30 | 2017-05-10 | 北京国双科技有限公司 | Data processing method and apparatus |
| CN107169854A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | A kind of method and device of data processing |
| CN107066318A (en) * | 2017-01-13 | 2017-08-18 | 广西电网有限责任公司电力科学研究院 | A kind of data dispatching method suitable for stream calculation |
| CN109800129A (en) * | 2019-01-17 | 2019-05-24 | 青岛特锐德电气股份有限公司 | A kind of real-time stream calculation monitoring system and method for processing monitoring big data |
| CN109634766A (en) * | 2019-02-20 | 2019-04-16 | 深圳大学 | Promote method, apparatus, equipment and the storage medium of data flow computer operational efficiency |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111767129A (en) | 2020-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8190864B1 (en) | APIC implementation for a highly-threaded x86 processor | |
| US10230608B2 (en) | RPS support for NFV by system call bypass | |
| US7386646B2 (en) | System and method for interrupt distribution in a multithread processor | |
| US7810096B2 (en) | Computer executing multiple operating systems | |
| US8219220B2 (en) | Industrial controller using shared memory multicore architecture | |
| US6425038B1 (en) | Conversion of desk-top operating system for real-time control using installable interrupt service routines | |
| US7047484B1 (en) | Method, system, and apparatus for providing access to asynchronous data in a spreadsheet application program | |
| US7412590B2 (en) | Information processing apparatus and context switching method | |
| US8019947B2 (en) | Technique for thread communication and synchronization | |
| US11360809B2 (en) | Multithreaded processor core with hardware-assisted task scheduling | |
| JPH09269903A (en) | Process managing system | |
| EP0614139A2 (en) | External procedure call for distributed processing environment | |
| CN111767129B (en) | Data flow task processing device and method | |
| KR930009633B1 (en) | Interrupt handling method for a data processor | |
| EP2630577B1 (en) | Exception control in a multiprocessor system | |
| JPH06324888A (en) | Scheduling system | |
| JPH0217541A (en) | Stack system micro-computer | |
| CN116166468A (en) | Method for processing ECC errors in heterogeneous system, heterogeneous system and related products thereof | |
| CN117234674A (en) | Method for performing task scheduling and related products | |
| US20150058505A1 (en) | Method for operating a buffer memory of a data processing system and data processing system | |
| CN117573386A (en) | Inter-process communication method and device | |
| JPS61201338A (en) | Processor for data flow | |
| JPH03100853A (en) | Inter-processor communication system | |
| KR940015818A (en) | How to Dispatch Processes in a Multiprocessing System | |
| JPH0495137A (en) | Context switching method for operating system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |