The devices library implements and provides a set of interfaces that you can use to interact with physical devices.
Library functions are used in backend functions to obtain information about devices. To select a device for executing the neural network, first call the backend function that provides computer types. The computer types are Device class inheritors with their own set of functions and properties. To select a device from those provided, call the select_devices() function by passing to it unique device IDs or pointers to Device instances.
You can obtain information about the device (for example, about power consumption) using the methods of the object that provides the interface to the device.
You can implement and use your own devices when working with source code if needed. To do this, implement a device class within the backend, provide an interface for the device, and implement a device loading function that returns a pointer to an instance of the Device class.
The Kaspersky Neuromorphic Platform currently has interfaces for the central processor and the AltAI-1 neuromorphic processing unit. When using an AltAI-1 neuromorphic processing unit, you can choose between a hardware processor and a software emulator as the device type. The hardware processor unit uses a fixed core configuration. If selecting a software emulator as the device type, you can change the configuration settings of the AltAI-1 neuromorphic processing unit cores by using the load_core_grid_params() method, and set the location of the input and output ports in the core configuration by using the make_standart_north_port() method.