公开(公告)号：AU2020274862A1

公开(公告)日：2021-10-14

申请号：AU2020274862

申请日：2020-05-12

Applicant: IBM

Inventor： LE GALLO-BOURDEAU MANUEL ,  KHADDAM-ALJAMEH RIDUAN ,  KULL LUKAS ,  FRANCESE PIER ANDREA ,  TOIFL THOMAS ,  SEBASTIAN ABU ,  ELEFTHERIOU EVANGELOS STAVROS

IPC: G06N3/063 ,  G06N3/08

Abstract: Methods and apparatus are provided for training an artificial neural network having a succession of neuron layers with interposed synaptic layers each having a respective set of

公开(公告)号：CA3137231A1

公开(公告)日：2020-11-19

申请号：CA3137231

申请日：2020-05-12

Applicant: IBM

Inventor： LE GALLO-BOURDEAU MANUEL ,  KHADDAM-ALJAMEH RIDUAN ,  KULL LUKAS ,  FRANCESE PIER ANDREA ,  TOIFL THOMAS ,  SEBASTIAN ABU ,  ELEFTHERIOU EVANGELOS STAVROS

IPC: G06N3/063 ,  G06N3/08

Abstract: Methods and apparatus are provided for training an artificial neural network having a succession of neuron layers with interposed synaptic layers each having a respective set of N-bit fixed-point weights {w} for weighting signals propagated between its adjacent neuron layers, via an iterative cycle of signal propagation and weight-update calculation operations. Such a method includes, for each synaptic layer, storing a plurality p of the least-significant bits of each N-bit weight w in digital memory, and storing the next n-bit portion of each weight w in an analog multiply-accumulate unit comprising an array of digital memory elements. Each digital memory element comprises n binary memory cells for storing respective bits of the n-bit portion of a weight, where n = 1 and (p + n + m) = N where m = 0 corresponds to a defined number of most-significant zero bits in weights of the synaptic layer.

公开(公告)号：IL288055D0

公开(公告)日：2022-01-01

申请号：IL28805521

申请日：2021-11-11

Applicant: IBM ,  LE GALLO BOURDEAU MANUEL ,  KHADDAM ALJAMEH RIDUAN ,  KULL LUKAS ,  FRANCESE PIER ANDREA ,  TOIFL THOMAS ,  SEBASTIAN ABU ,  ELEFTHERIOU EVANGELOS STAVROS

Inventor： LE GALLO-BOURDEAU MANUEL ,  KHADDAM-ALJAMEH RIDUAN ,  KULL LUKAS ,  FRANCESE PIER ANDREA ,  TOIFL THOMAS ,  SEBASTIAN ABU ,  ELEFTHERIOU EVANGELOS STAVROS

Abstract: Methods and apparatus are provided for training an artificial neural network having a succession of neuron layers with interposed synaptic layers each having a respective set of N-bit fixed-point weights {w} for weighting signals propagated between its adjacent neuron layers, via an iterative cycle of signal propagation and weight-update calculation operations. Such a method includes, for each synaptic layer, storing a plurality p of the least-significant bits of each N-bit weight w in digital memory, and storing the next n-bit portion of each weight w in an analog multiply-accumulate unit comprising an array of digital memory elements. Each digital memory element comprises n binary memory cells for storing respective bits of the n-bit portion of a weight, where n≥1 and (p+n+m)=N where m≥0 corresponds to a defined number of most-significant zero bits in weights of the synaptic layer.

公开(公告)号：SG11202110345XA

公开(公告)日：2021-10-28

申请号：SG11202110345X

申请日：2020-05-12

Applicant: IBM

Inventor： LE GALLO-BOURDEAU MANUEL ,  KHADDAM-ALJAMEH RIDUAN ,  KULL LUKAS ,  FRANCESE PIER ANDREA ,  TOIFL THOMAS ,  SEBASTIAN ABU ,  ELEFTHERIOU EVANGELOS STAVROS

IPC: G06N3/063 ,  G06N3/08

Abstract: Methods and apparatus are provided for training an artificial neural network having a succession of neuron layers with interposed synaptic layers each having a respective set of N-bit fixed-point weights {w} for weighting signals propagated between its adjacent neuron layers, via an iterative cycle of signal propagation and weight-update calculation operations. Such a method includes, for each synaptic layer, storing a plurality p of the least-significant bits of each N-bit weight w in digital memory, and storing the next n-bit portion of each weight w in an analog multiply-accumulate unit comprising an array of digital memory elements. Each digital memory element comprises n binary memory cells for storing respective bits of the n-bit portion of a weight, where n≥1 and (p+n+m)=N where m≥0 corresponds to a defined number of most-significant zero bits in weights of the synaptic layer.