From charlesreid1

Line 9: Line 9:
[[File:01_thumb.jpg|link=File:01_Updraft.pdf]]
[[File:01_thumb.jpg|link=File:01_Updraft.pdf]]


Link to the presentation PDF file: [[File:01_Updraft.pdf]]
'''Files:'''


Link to the presentation video file: [http://newton.endoftheinternet.org:8020/01_Updraft.mov Updraft Presentation Screencast]
* Presentation PDF file: [[File:01_Updraft.pdf]]


Wednesday, June 30, 2010
* Presentation video file [http://newton.endoftheinternet.org:8020/01_Updraft.mov Updraft Presentation Screencast] (direct link, right-click and pick "Save Link" to download)


Presenter: Charles Reid
'''Date:''' Wednesday, June 30, 2010
 
'''Presenter:''' Charles Reid


'''Abstract:''' One very important aspect of scientific computing is high-performance computing - running scientific software on machines with more than one processor (sometimes thousands!). This workshop will provide users with a clear understanding of cluster computing, and will focus specifically on the Updraft computing cluster, located at the University of Utah and administered by the Center for High Performance Computing. This will include a discussion of the hardware, software, policies, and queue system of the Updraft cluster, and an explanation of some important topics like head nodes vs. compute nodes, disk space and disk write speeds, and resources for finding the information you need. In addition, some useful Unix commands will also be covered.
'''Abstract:''' One very important aspect of scientific computing is high-performance computing - running scientific software on machines with more than one processor (sometimes thousands!). This workshop will provide users with a clear understanding of cluster computing, and will focus specifically on the Updraft computing cluster, located at the University of Utah and administered by the Center for High Performance Computing. This will include a discussion of the hardware, software, policies, and queue system of the Updraft cluster, and an explanation of some important topics like head nodes vs. compute nodes, disk space and disk write speeds, and resources for finding the information you need. In addition, some useful Unix commands will also be covered.

Revision as of 23:02, 1 October 2010


Research Presentations

2010 Scientific Computing Summer Workshop

Scientific Computing Summer Workshop 1: Updraft

01 thumb.jpg

Files:

Date: Wednesday, June 30, 2010

Presenter: Charles Reid

Abstract: One very important aspect of scientific computing is high-performance computing - running scientific software on machines with more than one processor (sometimes thousands!). This workshop will provide users with a clear understanding of cluster computing, and will focus specifically on the Updraft computing cluster, located at the University of Utah and administered by the Center for High Performance Computing. This will include a discussion of the hardware, software, policies, and queue system of the Updraft cluster, and an explanation of some important topics like head nodes vs. compute nodes, disk space and disk write speeds, and resources for finding the information you need. In addition, some useful Unix commands will also be covered.

Scientific Computing Summer Workshop 2: Creating, Building, and Using Scientific Software

02 thumb.jpg

Cmake thumb.jpg

Wednesday, July 7, 2010

Presenter: Charles Reid, Professor James Sutherland

Abstract: One extremely important but often overlooked aspect of scientific computing is the process of creating, building, and using scientific software. Sometimes this software is written by you or your group, other times it is written by someone you'll never meet. This workshop will cover some important techniques for building your own software project, some tools to manage your software projects as they grow in complexity, and commonly-used build tools that are used in most large-scale third-party software projects, such as autoconf, GNU make, and cmake. This discussion will cover concepts essential to building software (compilers, binaries, "make" and "cmake", libraries, linking, etc.), as well as give hands-on demonstrations in building several useful tools. Also, a brief discussion about basic source code management and documentation management (critical to any software project) will be given.

'APPLE USERS!' Please download and install a copy of Xcode Developer Tools BEFORE the workshop (it will take a while!). Download from Apple's webpage here: Apple: Download Xcode Developer Tools (go through the "Mac Dev Center" link).

'PLEASE NOTE!' A working knowledge of Unix will be assumed during this workshop, so if you aren't comfortable with a command line interface, be prepared to spend some time familiarizing yourself with Unix systems before you can utilize all the information given in this workshop.

You can also find some basic information about Unix systems at the following links:

Unix Command-Line Basics (less extensive) and

File:IntroToUnix.pdf - An Introduction to Unix (much more extensive).

However, everyone is encouraged to attend, regardless of their level of knowledge.

Scientific Computing Summer Workshop 3: Introduction to Arches and the Uintah Computational Framework

03 thumb.jpg

Wednesday, July 14, 2010

Presenter: Charles Reid

Abstract: The Arches large-eddy simulation code is an object-oriented C++ code developed to simulate turbulent combustion as a part of the C-SAFE program at the University of Utah. Arches is a component built within in the Uintah Computational Framework (UCF), a C++ framework enabling parallelization to thousands of processors. The Arches code is complex and multi-faceted, and as such requires the user to have a thorough understanding of both the UCF and the code's algorithm in order to begin developing within Arches. This workshop is intended to give potential Arches users and developers a gentle introduction to the UCF concepts required to understand what Arches is doing, as well as a higher-level algorithmic view of the Arches code.

It is recommended that users check out a copy of Uintah from SVN by running the following command at a command prompt:

$ svn co https://gforge.sci.utah.edu/svn/uintah/trunk ~/uintah

This will create a copy of the Uintah framework in a folder called "uintah" in your home directory. Checking out the code from the SVN repository will take a few minutes. The Arches component, if you are interested in exploring ahead of time, is located at the following location:

$ cd ~/uintah

$ cd src/CCA/Components/Arches 

Scientific Computing Summer Workshop 4: Stocking Your Unix Toolbox

04 thumb.jpg

Doxygen thumb.jpg

Wednesday, July 21, 2010

Presenter: Charles Reid

Abstract: In this workshop, we'll discuss some powerful Unix tools with which every engineer developing scientific software should have some familiarity. This will include a discussion of text editors (the classic "Emacs vs. Vi"), source code management and version-tracking software, software for documentation, and some tools provided by Unix that can prove themselves extremely useful and flexible.

The objective of the workshop is not to give detailed information about every tool; this would take much too long. Instead, the objective is to showcase some Unix tools to show their capabilities, and to give attendees a cursory introduction to a wide range of tools, so that the attendees can discover the tools that work best for them.

Stocking your Unix toolbox is something you must do yourself, but this workshop is intended to help get you started.

Supplementary Workshop: Using Star CCM+ on Updraft

Friday, July 2nd, 2010

Presenter: Michal Hradisky

Doctoral Program Presentations