CharmPy是一个通用的分布式并行编程框架，具有简单而强大的API，基于可迁移的Pytho对象和远程方法调用;构建于自适应C++运行时系统之上，提供速度、可扩展性和动态负载平衡。
CharmPy允许开发从笔记本电脑到超级计算机的并行应用程序。
简单示例：
from charmpy import charm, Chare, Group, Reducerfrom math import piimport timeclass Worker(Chare): def work(self, _steps, pi_future): h = 1.0 / _steps s = 0.0 for i i rage(self.thisIdex, _steps, charm.umPes()): x = h * (i + 0.5) s += 4.0 / (1.0 + x**2) # perform a reductio amog members of the group, sedig the result to the future self.cotribute(s * h, Reducer.sum, pi_future)def mai(args): _steps = 1000 if le(args) > 1: _steps = it(args[1]) mypi = charm.createFuture() workers = Group(Worker) # create oe istace of Worker o every processor t0 = time.time() workers.work(_steps, mypi) # ivoke 'work' method o every worker prit('Approximated value of pi is:', mypi.get(), # 'get' blocks util result arrives 'Error is', abs(mypi.get() - pi), 'Elapsed time=', time.time() - t0) charm.exit()charm.start(mai)
评论