Hardware analog-digital neural networks

A neural network comprising an analog-digital crosspoint-network, the analog-digital crosspoint-network comprising:

• a plurality of rows and columns;
• a plurality of synaptic nodes, each synaptic node of the plurality of synaptic nodes disposed at an intersection of a row and column of the plurality of rows and columns, wherein each synaptic node of the plurality of synaptic nodes includes a weight associated therewith;
• a column controller associated with each column of the plurality of columns, wherein each column controller is disposed to enable a weight change at a synaptic node in communication with that column controller;
• and
• a row controller associated with each row of the plurality of rows, wherein each row controller is disposed to control a weight change at a synaptic node in communication with that row controller;
• wherein both feedforward and feedback neural network matrices are determined through the network, and wherein a feedback weight matrix is the matrix transpose of a feedforward weight matrix.

A multistage neural network comprising a plurality of analog-digital crosspoint-networks, each network of the plurality of analog-digital crosspoint-networks comprising:

• a plurality of rows and columns;
• a plurality of synaptic nodes, each synaptic node of the plurality of synaptic nodes disposed at an intersection of a row and column of the plurality of rows and columns, wherein each synaptic node of the plurality of synaptic nodes includes a weight associated therewith;
• a column controller associated with each column of the plurality of columns, wherein each column controller is disposed to enable a weight change at a synaptic node in communication with that column controller;
• and
• a row controller associated with each row of the plurality of rows, wherein each row controller is disposed to control a weight change at a synaptic node in communication with that row controller;
• wherein a nonlinear column or row value from each stage of the multistage neural network is disposed as the row or column input value to an adjacent stage.