Convolutional neural network system employing resistance change memory cell array
Abstract:
A semiconductor integrated circuit includes a cell array, an input unit and an output unit. Cell array includes word lines, bit lines and resistance change cells respectively formed at cross points between word lines and bit lines. Input unit includes an access controller and a driver. Access controller controls access of data to a cell in time series, the data being expressed by a matrix. Driver applies voltage to a word line coupled to the cell which is an access destination of the data, the voltage being adjusted depending on a value of the data to be accessed to the cell. The output unit includes holding circuits each holding a representative value of an output level of a corresponding one of the bit lines in time series.
Public/Granted literature
Information query
Patent Agency Ranking
0/0