Selective deactivation of processing units for artificial neural networks

A method for operating an artificial neural network (ANN), the ANN including a consecutive series of layers, each of the layers mapping its input variables onto output variables, and, in each case the output variables of the layers are input variables of a particular layer that follows in the series, the ANN being implemented in a plurality of processing units in such a way that an implementation of each of the layers is split among at least two of the processing units, the method comprising the following steps:

• when a predefined deactivating condition is met during operation of the ANN independently of input variables supplied to it, at least one of the processing units is selectively deactivated in such a way that at least one further one of the processing units remains activated in all layers whose implementation is contributed to by the at least one of the processing units;
• and
• reactivating the at least one deactivated processing unit during operation of the ANN when a predefined activating condition is met, wherein:
• the layers map the input variables onto the output variables using a plurality of neurons, in the split of the implementation, computation of the plurality of neurons being split among at least two of the processing units,
• at least one of the at least two of the processing units has a share in the implementation of multiple layers of the consecutive layers, and
• by selectively deactivating the at least one of the processing units, the resettable switch-off device deactivates a subset of the plurality of neurons associated with the at least one of the processing units.

A non-transitory machine-readable storage device on which is stored a computer program that includes machine-readable instructions for operating an artificial neural network (ANN), the ANN including a consecutive series of layers, each of the layers mapping its input variables onto output variables, and, in each case the output variables of the layers are input variables of a particular layer that follows in the series, the ANN being implemented in a plurality of processing units in such a way that an implementation of each of the layers is split among at least two of the processing units, the computer program, when executed by a computer, causing the computer to perform the following steps:

• when a predefined deactivating condition is met during operation of the ANN independently of input variables supplied to it, at least one of the processing units is selectively deactivated in such a way that at least one further one of the processing units remains activated in all layers whose implementation is contributed to by the at least one of the processing units;
• and
• reactivating the at least one deactivated processing unit during operation of the ANN when a predefined activating condition is met, wherein:
• the layers map the input variables onto the output variables using a plurality of neurons, in the split of the implementation, computation of the plurality of neurons being split among at least two of the processing units,
• at least one of the at least two of the processing units has a share in the implementation of multiple layers of the consecutive layers, and
• by selectively deactivating the at least one of the processing units, the resettable switch-off device deactivates a subset of the plurality of neurons associated with the at least one of the processing units.