Course: Data Center Scale Computing Class

Project Details

Project Lead
Dirk Grunwald 
Project Manager
Dirk Grunwald 
Project Members
Sears Merritt, Eric Goodman, Matthew Monaco, Greg Guyles, Michael Coughlin, Andrew Sayler, Adam Cardenas, Oliver Michel, bo pearce, Abdullah Boodai  
Institution
Univ. of Colorado, Boulder, Computer Science  
Discipline
Computer Science (401) 

Abstract

This project is for a research seminar on data center scale computing and to support projects within that class. The class covers the technology, architecture and applications of data center scale computing. Projects involving computing-on-graphs, data-parallel processing and architectures similar to RAMcloud 

Intellectual Merit

This class has served to "boostrap" skills in data center scale computing and applications at the Univ. of Colorado. The course covers basic skills in using IaaS systems, services such as zookeeper and data-centric computing. This semester's class will also cover in-memory based computing, persistent memory systems and scalable graph processing.

Broader Impacts

This class serves graduate and some undergraduate students at the University of Colorado who are interested in applying data center scale computing to applications in graph processing, social network analysis and distributed systems.

Scale of Use

A few VM's for experiments / learning. We will be wanting to run benchmarks using hadoop for different graph systems, but that will be for smaller (8-16 nodes) systems.