README.rst

CharmPy

CharmPy is a general-purpose parallel and distributed programming framework with a simple and powerful API, based on migratable Python objects and remote method invocation; built on top of an adaptive C++ runtime system providing speed, scalability and dynamic load balancing.

CharmPy allows development of parallel applications that scale from laptops to supercomputers, using the Python language.

Please see the Documentation.

Short Example

The following computes Pi in parallel, using any number of machines and processors:

from charmpy import charm, Chare, Group, Reducer
from math import pi
import time

class Worker(Chare):

    def work(self, n_steps, pi_future):
        h = 1.0 / n_steps
        s = 0.0
        for i in range(self.thisIndex, n_steps, charm.numPes()):
            x = h * (i + 0.5)
            s += 4.0 / (1.0 + x**2)
        # perform a reduction among members of the group, sending the result to the future
        self.contribute(s * h, Reducer.sum, pi_future)

def main(args):
    n_steps = 1000
    if len(args) > 1:
        n_steps = int(args[1])
    mypi = charm.createFuture()
    workers = Group(Worker)  # create one instance of Worker on every processor
    t0 = time.time()
    workers.work(n_steps, mypi)  # invoke 'work' method on every worker
    print('Approximated value of pi is:', mypi.get(),  # 'get' blocks until result arrives
          'Error is', abs(mypi.get() - pi), 'Elapsed time=', time.time() - t0)
    charm.exit()

charm.start(main)

This is a simple example and demonstrates only a few features of CharmPy. Some things to note from this example:

• Chares are distributed Python objects.
• A Group is a type of distributed collection where one instance of the specified chare type is created on each processor.
• Remote method invocation in CharmPy is asynchronous.

In this example, there is only one chare per processor, but multiple chares (of the same or different type) can exist on any given processor, which can bring performance benefits. Please refer to the documentation for more information.

Contact

Mailing list

Main author at <[email protected]>