JPH06101046B2 - Data flow type information processing device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information processing device, and more particularly to executing a data flow program including instructions for performing a numerical operation process or a logical operation process on two data. The present invention relates to a data flow type information processing device.

[Prior Art] In a conventional Neumann computer, various instructions are stored in advance in a program memory as a program, and the addresses of the program memory are sequentially designated by a program counter so that the instructions are sequentially read and executed. To be done.

On the other hand, the data flow type information processing apparatus is a type of non-Neumann type computer that does not have the concept of sequential instruction execution by a program counter. Such a data flow type information processing apparatus employs an architecture based on parallel processing of instructions. In the data-flow type information processing device, instructions can be executed as soon as the data to be operated are available, and multiple instructions are driven simultaneously by the data, so the programs can be executed in parallel according to the natural flow of data. To be done. As a result, it is considered that the time required for calculation is significantly shortened.

FIG. 3 is a block diagram showing an example of the configuration of a conventional data flow type information processing apparatus. FIG. 4 is a diagram showing an example of the field structure of a data packet processed by the information processing apparatus.

The data packet 60 shown in FIG. 4 has a destination field.
41, instruction field 42, data 1 field 43, and data 2 field 44. Destination information 41 is stored in the destination field 41, instruction information is stored in the instruction field 42, and operand data is stored in the data 1 field 43 or the data 2 field 44.

The information processing apparatus shown in FIG. 3 includes a program storage unit 1, a pair data detection unit 2, and an arithmetic processing unit 3. A data flow program 70 shown in FIG. 5 is stored in the program storage unit 1. As shown in FIG. 5, the program storage unit 1 specifies the destination information 71 of the data flow program 70, the instruction information 72, and the copy presence / absence by the address designation based on the destination information of the input data packet 60.
The non-information 73 is read, the destination information 71 and the instruction information 72 are stored in the destination field 41 and the instruction field 42 of the data packet 60, and the data packet 60 is output. The paired data detection unit 2 waits for the data packet 60 output from the program storage unit 1. That is, two different data packets with the same destination information
60 is detected, and the operator land data of one of the data packets 60, for example, the contents of the data 1 field 43 in FIG. 4 is stored in the data 2 field 44 of the other data packet 60, and the other one is stored. The data packet 60 of is output. The arithmetic processing unit 3 performs arithmetic processing on the data packet 60 output from the paired data detection unit 2 based on the instruction information, stores the result in the data 1 field 43 of the data packet 60, and stores the result. Is output to the program storage unit 1.

The program storage unit 1 and the paired data detection unit 2 are connected by a data transmission line 4. Data transmission line 4 is 2
And is coupled to two input ports of the paired data detection unit 2. The data packet 60 output from the program storage unit 1 is selectively input to one of the paired data detection units 2 depending on whether the operand data is right operand data or left operand data in the arithmetic processing. Input to the port. Further, the paired data detection unit 2 and the arithmetic processing unit 3 are connected by a transmission line 5, and the arithmetic processing unit 3 and the program storage unit 1 are connected by a transmission line 6.

The data packet 60 continues to rotate in sequence through the program storage unit 1, the paired data detection unit 2, the arithmetic processing unit 3, and the program storage unit 1, so that the arithmetic processing based on the data flow program 70 stored in the program storage unit 1 is performed. proceed.

Consider the case where the data flow program 70 to be executed requires copying of data as shown in FIG. That is, when a certain operation result, for example, the output of the addition instruction in FIG. 6 is referred to by two or more other operations, for example, the multiplication and subtraction instructions in FIG. 6, copy processing is performed in the program storage unit 1. Be done. As shown in FIG. 5, copy presence / absence information 73 is stored in the program storage unit 1.

First, the data flow program 70 reads the contents of the portion addressed based on the destination information of the input data packet 60. At this time, with / without copy 73
If “NO”, the data packet 60 in which the contents of the destination field 41 and the instruction field 42 are updated is output and the process ends. On the other hand, copy Yes / No information 73 is “Yes”
In this case, the data packet 60 in which the contents of the destination field 41 and the instruction field 42 are updated is output, and the destination information 71, the instruction information 72 and the copy presence / absence information 73 that are stored subsequently are read out. . If the subsequently read copy presence / absence information 73 is “absent”, the same data as the input data packet 60 is the data 1 field.
Data packet 60 in which destination information 71 and instruction information 72 stored in 43 and currently read are stored in destination field 41 and instruction field 42, respectively, is output, and the process ends. If the copy presence / absence information 73 read out subsequently is “present”, similar copy processing is further continued.

In the above information processing apparatus, once the copy process is performed, the transmission line 6 for input to the program storage unit 1 and the transmission line 4 for output from the program storage unit 1
The flow rate ratio of the data packet 60 is always 1: 2.
Here, the flow rate ratio of the data packet 60 is the ratio of the number of data packets 60 passing through those transmission paths per unit time. FIG. 7 shows a data packet on the transmission line 4.
Data packet 60 of each transmission line when the flow rate of 60 is 1.
The flow rate of The paired data detection unit 2 outputs one data packet 60 corresponding to the input of two data packets 60. Therefore, even if it is assumed that the transmission line 4 operates at the maximum performance, only the flow rate of half the maximum performance is guaranteed after the transmission line 5.

In this way, when the copy processing is performed in the program storage unit 1, the transmission lines 5 and 6 have the maximum performance of 2
It only works at one-half the performance. Therefore, the arithmetic processing unit 3 is also capable of exerting only one half of the potential, and as a result, the performance is degraded when the information processing device executes the program.

FIG. 8 is a block diagram showing another example of the configuration of the conventional data flow type information processing apparatus.

The information processing apparatus shown in FIG. 8 includes two program storage units 11
And 12 are provided. The data packet 60 output from the arithmetic processing unit 3 is transferred to the program storage unit via the allocation unit 13.
Input to either 11 or 12. Arithmetic processing unit 3
The allocating unit 13 and the allocating unit 13 are connected by the transmission line 6. Allocation unit 13
Are coupled to program storage units 11 and 12 via transmission lines 16 and 17, respectively. In addition, the program storage unit 11
And 12 are individually coupled to the paired data detector 2 via transmission lines 14 and 15, respectively. The function of each part of the information processing apparatus in FIG. 8 is exactly the same as that of each part in the information processing apparatus in FIG. The allocation unit 13 distributes the data packet 60 output from the arithmetic processing unit 3 to one of the two program storage units 11 and 12 according to the arrival order.

FIG. 9 shows a data packet 60 on the transmission lines 14 and 15.
The flow rate of the data packet 60 on each transmission line is shown when the flow rate of 1 is set to 1. In the information processing apparatus of FIG. 8, the data packet 60 to the program storage units 11 and 12 is
Since the input of the above is distributed by the allocation unit 1, the data packets 60 are input to the program storage units 11 and 12 at a rate of half the maximum performance. When the copy processing is performed in the program storage units 11 and 12, the flow rate of the data packet 60 output from the program storage units 11 and 12 to the transmission lines 14 and 15 is the same as that of the program storage unit 11 and the transmission lines 16 and 17. Data packet 60 input to 12
And twice the flow rate. That is, in the transmission lines 14 and 15 for inputting the data packet 60 to the paired data detection unit 2, the maximum flow rate can be ensured. Also,
In the transmission line 5 for the data packet 60 output from the paired data detection unit 2, that is, the transmission line 5 for the data packet 60 input to the arithmetic processing unit 3, the maximum performance flow rate can be secured.

Therefore, in the conventional information processing apparatus shown in FIG.
The performance of the arithmetic processing unit 3 is maximized, and the effect of parallel processing is theoretically doubled as compared with the conventional information processing apparatus shown in FIG.

[Problems to be Solved by the Invention] In the conventional data flow type information processing apparatus, as shown in FIG. 8 described above, in order to improve the processing efficiency of the data flow type information processing apparatus shown in FIG. In addition, a data flow type information processing device having a dual program storage unit has become possible.

However, in the data flow type information processing apparatus shown in FIG. 8, since the program storage unit 1 is duplicated and provided, the total storage capacity of the program storage unit 1 is larger than that of the apparatus of FIG. Doubled. Therefore, there is a problem in that the information processing apparatus itself becomes large in scale.

Further, based on the execution order of the data flow program 70 stored in the program storage unit 1, the data packet 60 that sequentially makes a cycle between the processing units in the device is always the fourth packet described above.
Since the processing was performed in the form shown in the figure, the width of the data line (path) connecting the respective processing units was wasted. That is, originally, in reading the data flow program 70 from the program storage unit 1, only the data in the destination field 41 in the data packet 60 indicating the next processing instruction is necessary, and the instruction field 42 and the data 1 in the data packet 60 are required. The data in field 43 and data 2 field 44 were unnecessary. Also in the arithmetic processing unit 3, the instruction field 42, the data 1 field 43 and the data 2 in the data packet 60 are included.
Data in field 44 required, data packet 60
The data in destination field 41 in it was unnecessary. Therefore, there is also a problem that a large amount of waste is physically generated when configuring the device.

Therefore, an object of the present invention is to reduce the expansion of the physical scale of the device itself by constructing a data packet that circulates between the processing units in the device with the minimum data required for each processing unit. It is to provide such a data flow type information processing device.

[Means for Solving the Problem] A dataflow type information processing apparatus according to the present invention is a dataflow type information processing apparatus that performs information processing based on data including a portion required for calculation and a portion required for program reading. And a program storing means for storing the data flow program, an arithmetic processing means for performing arithmetic processing of data according to the data flow program, information read from the program storing means, and a result calculated by the arithmetic processing means. Merging means for coupling, paired data detection means for processing the data coupled by the confluent means, a part required for computation and an output for the paired data detection means And branching means for separating into

[Operation] In the information processing apparatus according to the present invention, a branching unit for newly adding identification information to separate data in the apparatus and the data separated by the branching unit are used as identification information added to the data. Since the merging portion which connects according to the above is provided and the minimum amount of data required by each processing unit is sent, the width of the data line connecting between the processing units can be reduced.

[Embodiment] An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing the configuration of a data flow type information processing apparatus according to an embodiment of the present invention.

In FIG. 1, the processing contents of the program storage unit 1, the paired data detection unit 2, and the arithmetic processing unit 3 are the same as those of the conventional information processing apparatus. In this embodiment, the identification tag separation unit 21 is located at the entrance of the paired data detection unit 2, and the identification tag attachment unit 22 is located at the exit of the paired data detection unit 2.

In this embodiment, the output of the program storage unit 1 and the output of the arithmetic processing unit 3 are connected to the two inputs of the identification tag separation unit 21, and the two outputs of the identification tag separation unit 21 of the paired data detection unit 2. It is connected to two inputs. Further, the output of the paired data detection unit 2 is connected to the input of the identification tag addition unit 22,
The two outputs of the identification tag adding section 22 are connected to the respective inputs of the program storage section 1 and the arithmetic processing section 3.

At this time, the program storage unit 1 and the identification tag separation unit 21 are connected by the transmission line 31, and the arithmetic processing unit 3 and the identification tag separation unit 21 are connected by the transmission line 32. The identification tag separation unit 21 and the paired data detection unit 2 are connected by two transmission lines, that is, a transmission line 33 and a transmission line 34, and the paired data detection unit 2 and the identification tag attachment unit 22 are connected by a transmission line 35. Has been done. Further, the identification tag attachment unit 22 and the program storage unit 1 are connected by a transmission line 36, and the identification tag attachment unit 22 and the arithmetic processing unit 3 are connected by a transmission line 37.

FIG. 2 is a diagram showing a field structure of a data packet processed in the information processing apparatus of FIG. Each data packet shown in FIG. 2 is generated by each processing unit shown in FIG. In FIG. 2, the data packet 61
Consists of an identification tag 50, an instruction field 42, a data 1 field 43 and a data 2 field 44. Identification tag 50
Is one of the marks when connecting two data packets. The data packet 62 consists of an identification tag 50 and a destination field 41. The data packet 63 consists of an identification tag 50 and a result data field 45. The result data feed 45 stores the calculation result performed by the calculation processing unit 3. The data packet 64 comprises an identification lag 50, a destination field 41 and a command field 42. The identification tag 50 is one of the marks when connecting two data packets, and the result data field 45 stores the calculation result performed by the calculation processing unit 3. The other fields store the same data as conventional ones.

Next, the processing operation of the information processing apparatus shown in FIG. 1 will be described with reference to FIG. 1, FIG. 2, FIG. 4, and FIG.

First, the data packet 60 is sent from the program storage unit 1 to the identification tag separation unit 21. In the data packet 60, data to be processed is stored in the data 1 field 43 or the data 2 field 44. At this time,
Since the identification tag 50 is not attached to the data packet 60, the identification tag separation unit 21 does not separate the identification tag 50 and sends it to the paired data detection unit 2 via the transmission path 33. The pair-to-data detection unit 2 waits for packets with matching destination fields 41, and when matching packets are prepared,
The data packet 60 is generated in the same manner as the conventional processing. Data packet 60 generated by the data detection unit 2
Is sent to the identification tag attachment unit 22 via the transmission path 35.

At this time, in the data packet 60, the designated address for reading the next instruction from the data flow program 70 stored in the program storage unit 1 is stored in the destination field 41. In addition, the operation contents of the
Data stored in
43 or a data 2 field 44 (a pair is used in a calculation that requires two data, and one is used in a calculation of one data).

The identification tag attachment unit 22 divides the data packet 60 into a data packet 61 and a data packet 62. Identification tag 50 attached to data packet 61 and data packet 62
Indicates that both packets are a pair generated from the same data packet 60. The data packet 62 generated by the identification tag attachment unit 22 is sent to the program storage unit 1 via the transmission line 36 in order to read the next command from the data flow program 70 stored in the program storage unit 1. To be Further, the data packet 61 is sent to the arithmetic processing unit 3 via the transmission line 37 in order to perform data arithmetic processing.

In the program storage unit 1, the data packet 64 is generated from the information of the data packet 62 sent from the identification tag attaching unit 22. That is, the destination information 71 of the data flow program 70 specified by the address stored in the destination field 41 of the data packet 62 is stored in the destination field 41 of the data packet 64. Further, the instruction information 72 of the data flow program 70 is stored in the instruction field 42 of the data packet 64. The identification tag 50 of the data packet 62 is added to the identification tag 50 of the data packet 64. The data packet 64 generated in the program storage unit 1 is sent to the identification tag separation unit 21 via the transmission line 31.

The arithmetic processing unit 3 generates a data packet 63 based on the information of the data packet 61 sent from the identification tag attaching unit 22. That is, using the data stored in the data 1 field 43 and the data packet 2 field 44 of the data packet 61, the command field of the data packet 61 is used.
Calculation processing is performed based on the calculation contents stored in 42. The calculation result is stored in the result data field 45 of the data packet 63. Identification tag 50 of data packet 61
Is attached to the identification tag 50 of the data packet 63. The data packet 63 generated by the arithmetic processing unit 3 is sent to the identification tag separating unit 21 via the transmission path 32.

In the identification tag separation unit 21, the data packet 64 sent from the program storage unit 1 and the data packet 63 sent from the arithmetic processing unit 3 join together.

In the identification tag separation unit 21, a data packet 60 is generated from the merged data packet 63 and data packet 64.
That is, one packet is generated with the identification tag 50 of the data packet 63 and the data packet 64 as a mark. The data packet 60 generated by the identification tag separation unit 21
Is sent to the paired data detection unit 2 via the transmission path 33 or the transmission path 34. After that, the process according to the data flow program 70 proceeds through each device in order.

Therefore, in this embodiment, since the data packet connecting only the data required by each processing unit is generated and transmitted, the data line through which the data packet passes is limited to the required width of the transmission data packet. Can be reduced.

In the embodiment described above, the identification tag separation unit 21
Waiting for a data packet with a matching 50 will be performed. Therefore, the identification tag separating unit 21 has an internal buffer or an external buffer (not shown) for waiting.

[Effects of the Invention] As described above, according to the present invention, a data packet passing between the processing units of the information processing apparatus is generated by connecting only the minimum data required for the processing by the processing units. However, since it is sent to the transmission path, there is an effect that the width of the data line connecting between the processing units can be reduced, and further, the area of the IC chip, which is a structural unit of each processing unit, can be reduced.

When the information processing apparatus has a plurality of processing units such as the program storage unit, the paired data detection unit, and the arithmetic processing unit, the above effects are effective in reducing the scale of the device itself.

Further, even when the information processing apparatus is configured as a multiprocessor by combining a plurality of units with the unit shown in FIG. 1 as one unit, the above effect is effective in reducing the scale of the apparatus itself.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram showing the configuration of a data flow type information processing apparatus according to an embodiment of the present invention. FIG. 2 is a diagram showing a field structure of a data packet processed in the information processing apparatus of FIG. FIG. 3 is a diagram showing an example of the configuration of a conventional data flow type information processing apparatus. Figure 4 shows
It is a figure which shows the field structure of the data packet processed in a data flow type information processing apparatus. Figure 5 shows
It is a figure which shows a part of data flow program stored in the program storage part of a data flow type information processing apparatus.
FIG. 6 is a diagram showing a part of a data flow program including a copy process. FIG. 7 is a diagram showing the flow rate of data packets in each transmission line of the information processing apparatus of FIG. FIG. 8 is a diagram showing another example of the configuration of the conventional data flow type information processing apparatus. FIG. 9 is a diagram showing the flow rate of data packets in each transmission line of the information processing apparatus of FIG. In the figure, 1 is a program storage unit, 2 is a pair data detection unit, 3
Is an arithmetic processing unit, 21 is an identification tag separating unit, 22 is an identification tag adding unit, 41 is a destination field, 42 is an instruction field, 43 is a data 1 field, 44 is a data 2 field, 45 is a result data field, and 50 is an identification. Tags and and 60, 61,
62, 63, and 64 are data packets. In each drawing, the same reference numerals indicate the same or corresponding parts.

Claims (1)

[Claims] 1. A data flow type information processing device for performing information processing based on data including a part required for calculation and a part required for program reading, comprising a program storage means for storing a data flow program, and a data flow. Arithmetic processing means for performing arithmetic processing of data according to a program, merging means for combining the information read from the program storage means, the result calculated by the arithmetic processing means, and the data combined by the merging means A data flow type information processing including a paired data detection means in a processable form and a branching means for separating an output from the paired data detection means into a portion required for calculation and a portion required for program reading. apparatus.