Back to results
US 2022/0101119 A1Published application
Artificial neural network training in memory
AssigneeMicron Technology, Inc.
InventorsSaideep Tiku, Poorna Kale
Life of this patent
Estimating the term…
Abstract
Apparatuses and methods can be related to implementing age-based network training. An artificial neural network (ANN) can be trained by introducing errors into the ANN. The errors and the quantity of errors introduced into the ANN can be based on age-based characteristics of the memory device.
Claims20
A method, comprising:
- accessing, from a memory array of a memory device, data representing elements of an artificial neural network model, the elements comprising a plurality of weights and a plurality of biases;
- introducing logical values at one or more locations of a bit string, the logical values representative of a number of errors that correspond to an expected bit-error rate of the memory device, during training of the artificial neural network model;
- and
- implementing the artificial neural network model in the memory device based at least in part on writing the introduced logic values of the bit string to the memory array.
1 more element
An apparatus, comprising:
- a memory array;
- a controller coupled to the memory array and configured to:
- access data representing elements of an artificial neural network model, the elements comprising a plurality of nodes stored in the memory array;
- modify the plurality of nodes of the artificial neural network model to introduce logical values at one or more locations of a bit string, the logical values representing a number of errors, into the artificial neural network model, wherein a quantity of the number of errors corresponds to an expected bit-error rate of the memory array;
- train the artificial neural network model including the number of errors;
- and
- implement the trained artificial neural network model in circuitry at the edge of the memory array based at least in part on writing the introduced logic values of the bit string to the memory array.
4 more elements
The apparatus of the 9, wherein the controller is further configured to determine a simulated age of the memory array.
A non-transitory computer-readable medium having computer-readable instructions, which when executed by a computer, cause the computer to:
- determine whether an error correction code (ECC) is utilized in execution of an artificial neural network;
- responsive to determining that the ECC is utilized in execution of an artificial neural network, modify the artificial neural network model implemented at an edge of a memory array based on characteristics of the ECC;
- and
- implement the artificial neural network model at the edge of the memory array.
1 more element
Dependencies are inferred from the claim text.