lebedov

#!/usr/bin/env python

"""
Demo of how to pass GPU memory managed by pycuda to mpi4py.

Notes
-----
This code can be used to perform peer-to-peer communication of data via
NVIDIA's GPUDirect technology if mpi4py has been built against a
CUDA-enabled MPI implementation.

codesections

use futures::{Async, Future, Poll};
use tokio::io::{AsyncRead, AsyncWrite, Error, ReadHalf, WriteHalf};
use tokio::net::TcpStream;
use warp::{path, Filter, Stream};

struct Receiver {
    rx: ReadHalf<TcpStream>,
}
impl Stream for Receiver {
    type Item = String;

FedericoPonzi

#!/bin/bash
#Federico Ponzi
#doylefermi
#chocolatkey
# GPLv2

usage()
{
    echo "Usage: $0 [-all][-list][-i] xxx.xxx.xxx.xxx"
}

funkyboy

# To pull the image
docker pull funkyboy/baasbox:0.9.2

# To run
docker run -d -p 9000:9000 -v /folder/on/host/machine:/baasbox/db funkyboy/baasbox:0.9.2