Closed sebastianmonten closed 1 month ago
currently the Nucleo-H743ZI2 board runs out of memory when input layer is 128x128 neurons. it is also too slow, see an example below:
Network shape: [16384, 3, 2] Simulation run until 10000 neuron updates reached Input frequency: 8600 spikes / second Neuron computations: 5250 / second
a network with 128x128 as input layer can be uploaded and run on the nucleo board
description:
currently the Nucleo-H743ZI2 board runs out of memory when input layer is 128x128 neurons. it is also too slow, see an example below:
dod:
a network with 128x128 as input layer can be uploaded and run on the nucleo board