Abstract: According to an embodiment, a neural network device includes: a plurality of cores each executing computation and processing of a partial component in a neural network; and a plurality of routers transmitting data output from each core to one of the plurality of cores such that computation and processing are executed according to structure of the neural network. Each of the plurality of cores outputs at least one of a forward data and a backward data propagated through the neural network in a forward direction and a backward direction, respectively. Each of the plurality of routers is included in one of a plurality of partial regions each being a forward region or a backward region. A router included in the forward region and a router included in the backward region transmit the forward data and the backward data to other routers in the same partial regions, respectively.

Abstract: According to one embodiment, a memory system including a key-value store containing key-value data as a pair of a key and a value corresponding to the key, includes an interface, a memory block, an address acquisition circuit and a controller. The interface receives a data write/read request or a request based on the key-value store. The memory block has a data area for storing data and a metadata table containing the key-value data. The address acquisition circuit acquires an address in response to input of the key. The controller executes the data write/read request for the memory block, and outputs the address acquired to the memory block and executes the request based on the key-value store. The controller outputs the value corresponding to the key via the interface.

Abstract: According to an embodiment, a synapse circuit includes: a buffer that changes an output signal to a second logical value at a timing when an input signal exceeds a first threshold level, in a case where the output signal has a first logical value in a first mode, and changes the output signal to the second logical value at a timing when the input signal exceeds a reference level lower than the first threshold level, in a case where the output signal has the first logical value in a second mode; an adjusting unit that adjusts the first threshold level depending on a stored coefficient; and a mode switching unit that operates the buffer in the first mode during a period in which an acquired spike is not generated, and operates the buffer in the second mode during a period in which the spike is generated.

Abstract: A calculation device includes: M coefficient storage units provided corresponding to the M coefficients, each of the M coefficient storage units including a positive-side coefficient and a negative-side coefficient representing a coefficient corresponding to a sign of a difference; M multiplication units provided corresponding to the M input values, each of the M multiplication units calculating a positive-side multiply value obtained by multiplying the positive-side coefficient included in the corresponding coefficient storage unit by a sign inverted according to the corresponding input value and a negative-side multiply value obtained by multiplying the negative-side coefficient included in the corresponding coefficient storage unit by a sign inverted according to the corresponding input value; and an output unit outputting an value according to a difference between a positive-side sum value obtained by summing the M positive-side multiplication values and a negative-side sum value obtained by summing the M

Abstract: According to an embodiment, a reinforcement learning system includes a memristor array in which each of a plurality of first direction lines corresponds to one of a plurality of states, and each of a plurality of second direction lines corresponds to one of a plurality of actions, a first voltage application unit that individually applies voltage to the first direction lines, a second voltage application unit that individually applies voltage to the second direction lines, a action decision circuit that decides action to be selected by an agent in a state corresponding to a first direction line to which a readout voltage is applied, a action storage unit that stores action selected by the agent in each state that can be caused in an environment, and a trace storage unit that stores a time at which the state is caused by action selected by the agent.

Abstract: A memcapacitor according to an embodiment includes a first electrode, a first dielectric layer provided on the first electrode, a plurality of variable resistance portions provided separately from each other on the first dielectric layer, a second dielectric layer provided on the first dielectric layer and between the variable resistance portions, and a second electrode provided on the variable resistance portions and the second dielectric layer. Each of the variable resistance portions is formed of a material that allows diffusion of metal atoms constituting the second electrode to inside of the variable resistance portion, and the second dielectric layer is formed of a material that prevents diffusion of the metal atoms constituting the second electrode to inside of the second dielectric layer.

Abstract: According to one embodiment, a storage device includes a plurality of memory nodes. Each of memory nodes includes a plurality of input ports, a plurality of output ports, a selector, a packet controller and a memory. The selector outputs a packet input to the input port to one of the output ports. The packet controller controls the selector. The memory stores data. The memory nodes are mutually connected at the input ports and the output ports. The memory node has an address that is determined by its physical position. The packet controller switches the output port that outputs the packet based on information including at least a destination address of the packet and an address of the memory node having the packet controller when receiving a packet that is not addressed to the memory node having the packet controller.

Abstract: According to an embodiment, a neural network device includes a control unit, and a matrix computation unit. The control unit causes a plurality of layers to execute a forward process of propagating a plurality of signal values in a forward direction, and a backward process of propagating a plurality of error values in a backward direction. The matrix computation unit performs computation on a plurality of values propagated in the plurality of layers. The matrix computation unit includes (m×n) multipliers, and an addition circuit. The (m×n) multipliers are provided in one-to-one correspondence with (m×n) coefficients included in a coefficient matrix of m rows and n columns. The addition circuit switches a pattern for adding values output from the respective (m×n) multipliers between the forward process and the backward process.

Abstract: According to an embodiment, a neural network device includes a plurality of cores, and a plurality of routers. Each of the plurality of routers includes an input circuit and an output circuit. Each of the plurality of cores transmits at least one of forward direction data propagating in the neural network in a forward direction and reverse direction data propagating in the neural network in a reverse direction. The input circuit receives the forward direction data and the reverse direction data from any one of the plurality of cores and the plurality of routers. The output circuit or the input circuit selectively deletes the reverse direction data stored based on a request signal for requesting reception of data.

Abstract: An authentication server according to embodiments performs statistical processing on a plurality of pieces of ID data acquired from an electronic device including a PUF circuit generating the pieces of ID data (S1052 to S1071), determines whether the plurality of pieces of ID data are physical random numbers based on a result of the statistical processing (S1072), and when the plurality of pieces of ID data are determined to be physical random numbers, recognizes the result of authentication of the electronic device as a success of authentication (S1073), and when the plurality of pieces of ID data are determined not to be physical random numbers, recognizes a result of authentication of the electronic device as a failure of authentication (S1074).

Abstract: According to an embodiment, an arithmetic device is configured to receive M input signals each representing a two-state value and M coefficients to output an output signal representing a two-state value. The device includes a positive-side current source, a negative-side current source, M cross switches, a coefficient memory unit, and a comparator. The positive-side current source is configured to output a first voltage corresponding to a value of 1/L of the current output from a positive-side terminal. The negative-side current source is configured to output a second voltage corresponding to a value of 1/L of the current output from a negative-side terminal. The memory unit includes M cells corresponding to the respective M coefficients. The comparator is configured to output an output signal having a value corresponding to a comparison result of the first voltage with the second voltage. Each M cell includes a first resistor and a second resistor.

Abstract: According to one embodiment, a storage device includes a plurality of memory nodes. Each of memory nodes includes a plurality of input ports, a plurality of output ports, a selector, a packet controller and a memory. The selector outputs a packet input to the input port to one of the output ports. The packet controller controls the selector. The memory stores data. The memory nodes are mutually connected at the input ports and the output ports. The memory node has an address that is determined by its physical position. The packet controller switches the output port that outputs the packet based on information including at least a destination address of the packet and an address of the memory node having the packet controller when receiving a packet that is not addressed to the memory node having the packet controller.

Abstract: According to one embodiment, a quantum annealing machine 1 includes a quantum bit array 21 which includes a plurality of cells (quantum bits) 211 respectively including a floating gate 105, and a controller 10 which executes writing of data in the plurality of cells 211, and temporally controls tunneling of an electric charge with respect to the floating gate 105.

Abstract: According to an embodiment, a memory system includes a memory and a computation unit. Into the memory, data are written. The memory stores therein multiple check matrices. Each of the check matrices is associated with the number of errors in the written data. The computation unit is configured to perform a first error correction on the written data by selectively using, from among the check matrices, a check matrix associated with the number of errors recognized in the written data.

Abstract: According to an embodiment, a decoding device includes a variable node processor, a check node processor, a first forwarder, and a second forwarder. The variable node processor is configured to perform variable node processing on variable nodes defined by a code and output first messages. The check node processor is configured to perform check node processing on check nodes defined by the code based on the first messages and output second messages. The first forwarder is configured to forward one or more first messages remaining after excluding messages to be forwarded to one or more check nodes corresponding to one or more of the second messages having been stored in ae storage, to the check nodes. The second forwarder is configured to forward the second messages to the variable nodes and forward the one or more of the second messages to the storage.

Abstract: An arithmetic control device according to an embodiment controls arithmetic operations using a memory chip. The memory chip includes a memory cell array and a controller. The memory cell array includes a plurality of memory cells. The controller is configured to control access to the memory cell array. The arithmetic control device includes a first writing unit and a first reading unit. The first writing unit requests the controller to write a first value to a second cell near a first cell of the memory cell array. The first reading unit requests the controller to read a second value from the first cell after the first value is written to the second cell.

Abstract: According to one embodiment, a storage device includes a plurality of memory nodes. Each of memory nodes includes a plurality of input ports, a plurality of output ports, a selector, a packet controller and a memory. The selector outputs a packet input to the input port to one of the output ports. The packet controller controls the selector. The memory stores data. The memory nodes are mutually connected at the input ports and the output ports. The memory node has an address that is determined by its physical position. The packet controller switches the output port that outputs the packet based on information including at least a destination address of the packet and an address of the memory node having the packet controller when receiving a packet that is not addressed to the memory node having the packet controller.

Abstract: According to one embodiment, a memory system including a key-value store containing key-value data as a pair of a key and a value corresponding to the key, includes an interface, a memory block, an address acquisition circuit and a controller. The interface receives a data write/read request or a request based on the key-value store. The memory block has a data area for storing data and a metadata table containing the key-value data. The address acquisition circuit acquires an address in response to input of the key. The controller executes the data write/read request for the memory block, and outputs the address acquired to the memory block and executes the request based on the key-value store. The controller outputs the value corresponding to the key via the interface.

Abstract: According to an emboediment, a multiplier accumurator includes a controller, a high-order multiplier, a high-order accumulator, a low-order multiplier, and an output unit. The controller is configured to designate each digit within a range of the most significant digit in a coefficient for an input value to a stop digit as a target digit. The high-order multiplier is configured to calculate a high-order multiplication value by multiplying the input value, and a value and a weight of the target digit. The high-order accumulator is configured to calculate a high-order accumulation value by accumulatively adding the high-order multiplication values for input values. The low-order multiplier is configured to calculate a low-order multiplication value by multiplying an input value and a value of a digit smaller than the stop digit. The output unit is configured to output a value determined based on whether the high-order accumulation value exceeds a boundary value.

Abstract: An individual identification device (1) according to embodiments may identify a storage device (100) including one or more memory chips (40). The device comprises a first storage (40), a region allocator (15), and a hardware fingerprint generator (12). The first storage may be configured to store write data. The region allocator may be configured to write the write data in a specific region in each memory chip. The hardware fingerprint generator may be configured to generate hardware fingerprint data based on mismatch bits between the write data and read data read out from the specific region in each memory chip.