ccoo Goto Github PK

3402_project_2

A combined MPI and OpenMP implementation of 1-dimesnional finite element method This project is a continuation from the first project. The aim of the project is to implement the program from the first project into a combined MPI-OpenMP framework. This is quite a common apparoach for exploiting both coarse and fine grain parallelism in programs. A problem is partitioned coarsely at the top level and finely within each individual part. The coarse level partitioning is done using MPI and the finer level partitioning is usually done on a multi-core machine or on a Graphics Processing Unit (GPU). The aim of the project is to partition the Finite Element Program into smaller parts and distribute these parts to different computers by using MPI. The computation on each part would now occur within the individual machines using the cores available on those machines. The coding in the project will be minimal, it is the case quite often that parallelizing a piece of sequential code requires only small but well thought out modifications. The deliverables: The first deliverable is of course your modified C or FORTRAN code with appropriate MPI and OpenMP directives. You should also comment your code suitably, so that the code itself can be read and your modifications can be understood. You have to do a scalability study by using 2,4,6 and 8 machines in the IBM cluster. Each machine in the cluster has 4 cores. For each of the above number of machines, you should use 4 threads in each machine. Also you should do the above study using at least two problem sizes. The IBM machines each have 8 GB of RAM, so you should be able to allocate quite large arrays dynamically. You should plot the results in graphs and include in the report. You have to submit a document where you should explain how you have implemented the parallelism in the code and why. Include also all the decisions related to your implementation in the first project, so that the document can be read without checking your first project again. The document should also include the graphs mentioned above. The document should be strictly in pdf format. You should develop your code on the IBM cluster. Please make as much use of it as you can. However, I will make the login restricted after about ten days, so that you can conduct your final performance analysis without any interference. I will allocate slots for group members when only they can login. Of course the output from your parallelized program should match the output of the sequential program. Deadline: The submission deadline is 11:59 pm on November 4, through cssubmit. Note: The project can be done either individually or in a group consisting of a maximum of two students. Note: I will give you hints within a few days if there are requests for hints. You are also welcome to post your thoughts/designs, however, you should not post any code.

a3c-distributed_tensorflow

Distributed Tensorflow Implementation of Asynchronous Methods for Deep Reinforcement Learning

alien-stresser

HTTP flooder multi-threaded in python

analytics-zoo

Distributed Tensorflow, Keras and PyTorch on Apache Spark/Flink & Ray

atomicp2p

A high-availability peer-to-peer framework which base on multi-process and threading with distribute and DNS syncing mechanism

autogluon

AutoGluon: AutoML for Image, Text, and Tabular Data

awesome-blockchain-cn

收集所有区块链(BlockChain)技术开发相关资料，包括Fabric和Ethereum开发资料

awesome-hacking

A collection of various awesome lists for hackers, pentesters and security researchers

awesome-python

A curated list of awesome Python frameworks, libraries, software and resources

black_jack

21 points(21点) / 强化学习 / Black jack(黑杰克) / 蒙特卡洛法

blockchain

黑马程序员 120天全栈区块链开发 开源教程

blockchain-1

区块链 - 中文资源

blockchainsimpledemo

基于Python3区块链简单Demo，帮助了解区块链。

branch-and-bound-com-gurobi

Branch and Bound utilizando o Gurobi

branch-and-bound_parallel

Project of multi-core programming

branch-and-price-gurobi

Repository contains implementation of Branch and Prive for classical General Assignment Problem problem using Python and Gurobi solver.

branch_bound_solve

An introduction to the branch and bound method

branchandpricevrp

Test on how to implement Branch and Price for VRP with Gurobi (work in progress)

cg_parallel-machine-sched

This repo aim is to reproduce a column generation approach based on "Parallel machine scheduling by column generation" paper

deep-q-learning-paper-to-code

deep-reinforcement-learning-algorithms

31 projects in the framework of Deep Reinforcement Learning algorithms: Q-learning, DQN, PPO, DDPG, TD3, SAC, A2C and others. Each project is provided with a detailed training log.

deep-reinforcement-learning-for-dynamic-spectrum-access

Using multi-agent Deep Q Learning with LSTM cells (DRQN) to train multiple users in cognitive radio to learn to share scarce resource (channels) equally without communication

dist-keras

Distributed Deep Learning, with a focus on distributed training, using Keras and Apache Spark.

distributed_reinforcement_learning

implementation of distributed reinforcement learning with distributed tensorflow

dqn-multi-agent-rl

Deep Q-learning (DQN) for Multi-agent Reinforcement Learning (RL)

free-programming-books

:books: Freely available programming books

free-programming-books-zh_cn

:books: 免费的计算机编程类中文书籍，欢迎投稿

fucking-algorithm

刷算法全靠套路，认准 labuladong 就够了！English version supported! Crack LeetCode, not only how, but also why.

fucking-algorithm-book

《labuladong的算法小抄》顺序阅读版

gopslpnlpbb

Global Optimization for Pump Scheduling by LP-NLP Branch-and-Bound Gratien Bonvin, Sophie Demassey, Andrea Lodi (2018) Pump scheduling in drinking water distribution networks with an LP/NLP-based branch and bound. https://hal-mines-paristech.archives-ouvertes.fr/hal-02158535/