摘要:
A parallel processor system which includes a plurality of processors each for executing at least one of a plurality of mutually associated programs and a transfer circuit. The transfer circuit is connected to the processors, and is provided for transferring the data outputted from any one of the programs during execution of one program by any one of the processors to other processors to which a receiving program is allotted. The transfer operation is performed in response to a program identification code outputted during execution of the one program by one processor to identify the receiving program.
摘要:
A parallel processor system includes a transfer circuit and a plurality of processors each of which executes at least one of a plurality of mutually associated programs. The transfer circuit transfers data from a sending program allotted to one processor to a receiving program allotted to another processor by identifying the other processor and the receiving program based on a job number and within-job process number outputted by the sending program.
摘要:
A parallel processor system having a plurality of processor elements includes transfer information generation circuit for generating transfer information by adding to vector data a data identifier for the vector data and a destination processor element number, transmission circuit for sending the transfer information to a data communication path, receive circuit for holding the transfer information sent from the data communication path, and vector register for continuously reading related element data from the receive circuit based on the data identifiers generated by the transfer information generation circuit.
摘要:
A computer system having a plurality of processors assigned first and second address portions are connected to a plurality of switch circuits. A first group transfer networks are connected to a corresponding first group of the plurality of switch circuits. Each of the transfer networks concurrently transfer data among the switch circuits. The switch circuits are provided to processors of a first kind arranged in a plurality of processor groups. The processor groups of the first kind include processors with different values for first address portions and the same value for second address portions. Additional transfer networks, processors and switches functioning in a similar manner are provided to expand the above system. In another embodiment of the present invention a data transfer network is provided having a plurality of processors for data transfer. The network includes a plurality of multistage switches each belonging to one of plural stages and connected to the switches of a preceding stage and to switches of the succeeding stage. Each of the switches are arranged to receive packets from a preceding switch. A packet includes a target process address and data to be transferred. A path select device is connected to receive packets and is also connected to plural switches belonging to a next stage for the transfer of the received partial addresses and partial data. A control device is connected to receive the partial addresses and partial data and is responsive to a predetermined bit within the received partial addresses. The control means is responsive to the arrival of the first partial address of the packet.
摘要:
A computer comprising a circuit for writing a group of ordered data elements onto the main storage; a circuit for reading said group of data from the main storage; and a circuit which is connected to the writing circuit and to the reading circuit, and which ensures the sequence of main storage references between said writing circuit and said reading circuit such that said reading circuit will not read the data elements that have not yet been written by said writing circuit among said group of data elements.
摘要:
A parallel computer system includes a plurality of processors, each of which is placed in data communication with an interconnecting network. Pairs of a data signal and a data identification code, predetermined for the data signal, are received by each processor and stored in a memory. Structure is provided for reading a data signal belonging to one of the pairs having a data identification code designated by a data readout instruction.
摘要:
In a multiprocessor digital computer system ID data, coupled with data for which inter-processor communication is desired, is communicated from one processor and held temporarily with data in a receiver buffer (associative memory) in a receiving processor. This ID is divided into main ID data MK and sub ID data SK. Main ID data MK is used for searching data from a receive buffer. The sub ID data SK are used as an ID of the data in the receive processor.
摘要:
A plurality of elemental processors each include a local memory for storing data and task programs and an execution section for executing the task programs. A communications section transfers data among the processors. In a method of parallel processing with these elemental processors, a task program is executed in one of the processors. A detection operation is conducted to determine whether the data from the task program is to be copied to the local memories of other processors. The detection is based on predetermined information which is stored in the local memory of the processor which performs the task program and indicates which of the other processors will need the data. The detection also determines which of the other processors that will require access to the data are ready to receive the data. Concurrently and in parallel with the execution of the task program, the data is transmitted to each of the elemental processors that are awaiting the data with only a write instruction and without any communications instruction.
摘要:
In accessing a memory, each element processor executes a program constructed so as to designate an address belonging to a predetermined local address area for each element processor. When a memory write instruction is executed by an element processor, it is detected if the memory address designated by the instruction coincides with a predetermined address. If detected, a predetermined address belonging to a local address space of another element processor and assigned to the first-mentioned predetermined address, and the data written in response to the write instruction, are sent to the other element processor to instruct the data to be written therein as a copy data. A next task to be executed is decided independently for each element processor.
摘要:
In a data transfer network of the present invention, each switch is designed such that when the partial address necessary for a given switch to determine another switch belonging to the succeeding stage, to which a packet is to be delivered from the given switch, is included in the first one of plural subpackets supplied to the given switch and each having the partial address, the given switch starts its switching operation upon arrival of the first subpacket. In a preferred embodiment, when the partial address necessary for the succeeding switch to make its switching operation is not included in the first subpacket, the partial addresses are exchanged between the subpackets by the preceding switch so that the said partial address is now included in the first subpacket.