Abstract: Individual distributed processing nodes packetize distributed data for each weight of a neural network of a learning object in an order of a number of the weight, transmit the distributed data to an aggregation processing node, acquire aggregation data transmitted from the node in order, and update the weight of the neural network. The node acquires the transmitted distributed data, packetizes the aggregation data for which the distributed data of all the distributed processing nodes is aggregated for each weight, and transmits the aggregation data to the individual nodes. The individual nodes monitor an unreceived data amount which is a difference between data amounts of the transmitted distributed data and the acquired aggregation data, and when the unreceived data amount becomes equal to or larger than a threshold Ma, stops transmission of the distributed data until the unreceived data amount becomes equal to or smaller than a threshold Mb (Mb<Ma).

Abstract: A distributed deep learning system that can achieve speeding-up by processing learning in parallel at a large number of learning nodes connected with a communication network and perform faster cooperative processing among the learning nodes connected through the communication network is provided. The distributed deep learning system includes: a plurality of computing interconnect devices 1 connected with each other through a ring communication network 3 through which communication is possible in one direction; and a plurality of learning nodes 2 connected with the respective computing interconnect devices 1 in a one-to-one relation, and each computing interconnect device 1 executes communication packet transmission-reception processing between the learning nodes 2 and All-reduce processing simultaneously in parallel.

Abstract: A first distributed processing node sets, as intermediate aggregated data, distributed data generated by the own node and transmits this data to the distributed processing node having the next number designated in advance. The intermediate distributed processing node excluding the first and last distributed processing nodes calculates, for each of weights corresponding thereto, a sum of the received intermediate aggregated data and distributed data generated by the own node, generates intermediate aggregated data after update, and transmits this data to the distributed processing node having the next number designated in advance. The last distributed processing node calculates, for each of the weights corresponding thereto, a sum of the received intermediate aggregated data and distributed data generated by the own node, generates aggregated data, and transmits this data to the first and intermediate distributed processing nodes.

Abstract: An OLT (10) is provided with a priority control bypass circuit (16) and an encryption/decryption bypass circuit (17), or an ONU (20) is provided with a priority control bypass circuit (26) and an encryption/decryption bypass circuit (27), and one or both of encryption/decryption processing and priority control processing are bypassed in accordance with a priority control bypass instruction (BP) and an encryption/decryption bypass instruction (BE), which are set in advance. This reduces a processing delay that occurs in the OLT or the ONU.

Abstract: Each of distributed processing nodes [n] (n=1, . . . , and N) packetizes pieces of distributed data [m, n] as packets for every M weights w [m] ((m=1, . . . , and M) of a neural network to be learned in an order of numbers m, transmits the packets to a consolidation processing node, receives a packet transmitted from the consolidation processing node to acquire consolidated data R [m] in the order of numbers m and update the weights w [m] of the neural network on the basis of the consolidated data R [m].

Abstract: Each of learning nodes calculates a gradient of a loss function from an output result obtained when learning data is input to a neural network to be learned, generates a packet for a plurality of gradient components, and transmits the packet to the computing interconnect device. The computing interconnect device acquires the values of a plurality of gradient components stored in the packet transmitted from each of the learning nodes, performs a calculation process in which configuration values of gradients with respect to the same configuration parameter of the neural network are input on each of a plurality of configuration values of each gradient in parallel, generates a packet for the calculation results, and transmits the packet to each of the learning nodes. Each of the learning nodes updates the configuration parameters of the neural network based on the value stored in the packet.

Abstract: Each of learning nodes calculates gradients of a loss function from an output result obtained by inputting learning data to a learning target neural network, converts a calculation result into a packet, and transmits the packet to a computing interconnect device. The computing interconnect device receives the packet transmitted from each of the learning nodes, acquires a value of the gradients stored in the packet, calculates a sum of the gradients, converts a calculation result into a packet, and transmits the packet to each of the learning nodes. Each of the learning nodes receives the packet transmitted from the computing interconnect device and updates a constituent parameter of a neural network based on a value stored in the packet.

Abstract: A method, a computing system and a computer program product are provided. A link for use by a user to access a file is created. Content of the file is encrypted using a common key. The common key is encrypted using a public key of the user and is registered in the link. Access rights regarding the file are set for the user and registered in the link. The link includes information for use by the user to access the file when the access rights indicate that the user is authorized to access the file.

Abstract: A technique of operating a system is provided that processes data with one or more data processing modules provided in parallel. A synchronization token is input into at least one data processing module that is in an operational state from the one or more data processing modules provided in parallel, in response to a request to change allocation of the data. The allocation of the data is changed to the one or more data processing modules provided in parallel, after the synchronization token is input. In response to the synchronization token having arrived at a data processing module at a later stage, the at least one data processing module, that stopped the processing, starts processing data after the synchronization token is input to the at least one data processing module.

Abstract: There is provided with a lighting apparatus. An elongated first light emission unit and an elongated second light emission unit each extend in a longer side direction and a shorter side direction. The first light emission unit and the second light emission unit have respective end portions in the longer side direction that are connected to each other via a restricting mechanism having a shape that restricts relative movement of the first and second light emission units in the shorter side direction.

Abstract: A vehicle transmission includes a shaft; speed-changing gears; switching mechanisms; and a shifting mechanism. The shifting mechanism is provided with a double-meshing preventing mechanism configured to switch between a one-way state in which the switching mechanisms are hindered from moving in a downshift direction and allowed to move in an upshift direction and a free state in which the switching mechanisms are allowed to move in both the downshift direction and the upshift direction.

Abstract: The present invention may be a method, a computer system, and a computer program product. An embodiment of the present invention provides a method for judging data consistency in a database. In one embodiment, the method comprises the following: generating a property of data obtained from a first database; associating the property with an attribute of a data model to generate a data property definition; judging whether data obtained from a second database satisfies the data property definition or not; and outputting a result of the judgment. In another embodiment, the method comprises the following: generating a property of data obtained from a database; associating the property with an attribute of a data model to generate a data property definition; judging whether data which is stored in the database satisfies the data property definition or not; and outputting a result of the judgment.

Abstract: The present invention may be a method, a computer system, and a computer program product. An embodiment of the present invention provides a method for finding a problem in procedures described in a guide document for install and configuration of software. The method comprises calculating, using a dynamic programming matching, a distance between an install-and-configuration log generated by executing the install and configuration of the software according to the guide document at a user-side computer and a log template generated by executing the install and configuration of the software according to the guide document at an administrator-side computer, and finding a problem in the procedures, using the distance.

Abstract: A vehicle control apparatus includes an inverter controller. The inverter controller holds a plurality of control maps for an inverter. The inverter supplies electric power to a drive motor. The drive motor drives a drive wheel of the vehicle. The inverter controller selects any one of the plurality of control maps on a basis of a notification instruction that instructs to notify a driver of information. The inverter controller controls an operation of the inverter on a basis of the control map selected.

Abstract: A line sensor apparatus comprises a rod lens array in which a plurality of rod lenses are arrayed in a first direction, a line sensor substrate including a line sensor which receives light from the rod lens array, a pair of inner supporting members configured to support the line sensor substrate, a transmitter substrate including a transmitter which transfers data based on the light detected by the line sensor, and an outer supporting member configured to support the pair of inner supporting members and the transmitter substrate, wherein the pair of inner supporting members are arranged on two side surfaces of the rod lens array in a second direction perpendicular to the first direction.

Abstract: The ultrasound probe includes a piezoelectric element including a piezoelectric composition and an electrode that applies a voltage to the piezoelectric composition. The piezoelectric composition has piezoelectric characteristics expressed by any coordinates included in a region formed by a polyhedron having a plurality of predetermined points as vertexes in Cartesian coordinates (keff, ?33S, Ec) including variables keff, ?33S and Ec.

Abstract: A technique of operating a system is provided that processes data with one or more data processing modules provided in parallel. A synchronization token is input into at least one data processing module that is in an operational state from the one or more data processing modules provided in parallel, in response to a request to change allocation of the data. The allocation of the data is changed to the one or more data processing modules provided in parallel, after the synchronization token is input. In response to the synchronization token having arrived at a data processing module at a later stage, the at least one data processing module, that stopped the processing, starts processing data after the synchronization token is input to the at least one data processing module.

Abstract: An upstream allocation circuit (14) and a downstream allocation circuit (15) are provided in an OLT (1). For example, a superimposed frame obtained by bundling upstream frames (upstream control frames+upstream data frames) from all ONUS is input to the upstream allocation circuit (14) via a frame reproduction circuit (12-1). The superimposed frame may be generated at the stage of optical signals or generated after converting optical signals into electrical signals. The upstream allocation circuit (14) allocates each of the upstream control frames bundled into the superimposed frame to a predetermined PON control circuit (13) based on information (PON port number or LLID) added to the frames. The downstream allocation circuit (15) allocates, to a preset frame reproduction circuit (12), each downstream control frames output from the PON control circuits (13).

Abstract: An OLT (10) is provided with a priority control bypass circuit (16) and an encryption/decryption bypass circuit (17), or an ONU (20) is provided with a priority control bypass circuit (26) and an encryption/decryption bypass circuit (27), and one or both of encryption/decryption processing and priority control processing are bypassed in accordance with a priority control bypass instruction (BP) and an encryption/decryption bypass instruction (BE), which are set in advance. This reduces a processing delay that occurs in the OLT or the ONU.

Abstract: A solid electrolytic capacitor includes: a capacitor element including a sintered compact of a valve action metal, a dielectric layer, an electrolyte layer, and a cathode layer sequentially formed over a surface of the sintered compact, and an anode wire drawn out of the sintered compact; an anode terminal; a cathode terminal; and an exterior resin. The anode terminal includes a mounting portion and an upright portion. The upright portion has a trapezoidal shape. A length of a welding surface corresponding to an edge of the upright portion on the anode wire side is set to be longer than a length of the upright portion on the mounting portion side and to be longer than a width of the anode wire.