neurokernel.core_gpu.Manager

class neurokernel.core_gpu.Manager(required_args=['sel', 'sel_in', 'sel_out', 'sel_gpot', 'sel_spike'], ctrl_tag=1)

Module manager.

Instantiates, connects, starts, and stops modules comprised by an emulation. All modules and connections must be added to a module manager instance before they can be run.

ctrl_tag

int – MPI tag to identify control messages.

modules

dict – Module instances. Keyed by module object ID.

routing_table

routing_table.RoutingTable – Table of data transmission connections between modules.

rank_to_id

bidict.bidict – Mapping between MPI ranks and module object IDs.

__init__(required_args=['sel', 'sel_in', 'sel_out', 'sel_gpot', 'sel_spike'], ctrl_tag=1)

x.__init__(…) initializes x; see help(type(x)) for signature

Methods

__init__([required_args, ctrl_tag])	x.__init__(…) initializes x; see help(type(x)) for signature
add(target, id, *args, **kwargs)	Add a module class to the emulation.
connect(id_0, id_1, pat[, int_0, int_1])	Specify connection between two module instances with a Pattern instance.
process_worker_msg(msg)	Process the specified deserialized message from a worker.
quit()	Tell the workers to quit.
recv([tag])	Receive data from child process.
send(data, dest[, tag])	Send data to child process.
spawn()	Spawn MPI processes for and execute each of the managed targets.
start([steps])	Tell the workers to start processing data.
stop()	Tell the workers to stop processing data.
validate_args(target)	Check whether a class’ constructor has specific arguments.
wait()	Wait for execution to complete.

Attributes

average_step_sync_time	Average step synchronization time.
average_throughput	Average received data throughput per step.
intercomm	Intercommunicator to spawned processes.
log_on	Logger switch.
total_throughput	Total received data throughput.