Ohio High School student discovers 'world of supercomputing'

Kochs interns with OSC staff to analyze uses of OpenFOAM software

Austin Kochs, a third-year student at The Metro Early College High School, delivered a final presentation last week on his internship experiences at the Ohio Supercomputer Center during the winter quarter.

Kochs gave his presentation to teachers, parents and Ohio Supercomputer Center staff and offered a comparative review of OpenFOAM, an open-source computational fluid dynamics program used by industry and academic researchers. Dr. Steve Gordon, Austin Kochs, and Jim Giuliani

Kochs has been the seventh student from the Metro High School internship program to conduct research projects at OSC. Metro High School, a collaboration among Battelle, The Ohio State University and Franklin County Education Council, engages students in personalized math, science and technology (STEM) education. As part of their curriculum, third- and fourth-year students may participate in the College Access program of learning laboratories within the central Ohio area that includes an internship at an organization of interest.

“Working closely with Metro High school student interns allows us to delve into real-world applications that expand their skill sets beyond the classroom,” said Jim Giuliani, client and technology support manager at OSC.  “Austin is a good example of this program’s mutual advantages – we benefit from the students’ in-depth analysis, and they get a stronger idea of fields to pursue in college.”

The following are Kochs’ blog entries on the OSC staff pages about what he learned. You may review his final presentation by clicking here.

January 19, 2010 - 8:50 a.m. 
Well, here goes my first Blog entry over my time so far at the Ohio Supercomputer Center.

I have been working on trying to compile the software package OpenFOAM (a computational fluid dynamics modeling application) to run on the Glenn cluster with Jim Giuliani. We ran into some major complications early on when trying to use the Intel C Compiler, figured out how to solve them, and then tried again only to find that we didn’t have the right modules loaded. Eventually we decided to switch over the GNU C Compiler and, supposedly have successfully compiled a working version of version 1.6 of the software. (I’ll be testing our build later this week so we can get some performance values and begin to debug and optimize the program.)

Last week I conducted two interviews: one with Kathryn Kelley, Director of the Outreach Program, and one with Doug Johnson, Senior Systems Engineer of HPC Operations. I asked them questions such as “How does what you do relate to the mission of the OSC” and “What did you have to do in order to get to where you are now.” I will be incorporating their responses into a paper for a school assignment.

Well, that’s how my time has been spent thus far here at the Ohio Supercomputer Center (in a nutshell). I really like being here and helping out (a big thank you to Jim specifically for taking the time to allow me to work with him and to Dr. Gordon for being my mentor and allowing me to be a part here) and I can’t wait for what this week and the next weeks will bring.

January 22, 2010 - 9:41 a.m. 
Well, after a log and hard thought session I have decided to keep my posts with the snazzy title of “Weekly Blog Entry #.”

The compilation of OpenFOAM was a success, and I was able to begin testing it. Jim and I ran into some problems with certain required libraries not being accessible by normal users, however that was just a permissions error. We also found a problem with the paraFoam package and are trying to work that out. I think it happened because the machine tried to use two different versions of the package qmake to make paraFoam, however I guess I’ll find out.

I feel sort of taken over by OpenFOAM, now wanting to compile it on my new desktop machine to give it a real computational test. Oh well, such is my life.

My Internship advisor, Mr. Bluel, will be here on Tuesday. I’m hoping that he likes whatever it is that he’s looking for in my progress.

Have a good weekend, everyone!


January 29, 2010 - 9:29 a.m. 
This week was another productive week for OpenFOAM, although it’s not quite done yet. Early on, we ran into a problem with paraFoam not being able to find the _ZN11QHeaderView5resetEv symbol. I found that the problem may have been caused by accidentally using two different versions of qmake (provided by the Qt software package) during the compilation of paraview. It is currently being recompiled and hopefully we’ll have a fully functional OpenFOAM-1.6 installation come later today and begin testing on Monday.

I was able to learn more as to how OpenFOAM using multiple processors and the performance benefits (or negatives) that can come as a result. I now have this urge to build a five computer cluster at my house, just so I can get some more experience as to how to use, design, and administer a cluster (although mine would probably suffer from severe communications lag because I’d be stuck using gigabit ethernet cables to connect them).

On a more-or-less random note, I received an email from tidbits with a request for helping out Professor Passino (if I remember correctly, he is a part of the electrical and computer engineering department) with setting up some XO laptops as a part of the One Laptop Per Child project for him to take to an all-girls school/orphanage in Columbia. I’m so excited to help out. (I first learned about the One Laptop Per Child project maybe a year and a half ago and wanted to try and help out with it at some point in time.)

My final day here will be sometime around February 22nd, so I hope that the rest of my time here can be well spent.


 February 8, 2010 - 8:18 a.m. 
This week, I submitted my first actual job (not just compiling) to the cluster! I had been hoping to have something fun and visual (rather than just terminal commands) to show my adviser and a fellow student who is in charge of filming those of us out at sites, and on Thursday, we were ready with a coll engine simulation. Sadly they had to reschedule the meeting to Monday, but that just means I get a bit of time to figure out what sorts of cool things I can do with the simulation.

I’ve been having some troubles with ParaView, the OpenFOAM data visualizer as I wish to call it. After getting Xwin32 installed on my laptop I was finally able to run ParaView, however its glory was short lived. I can’t actually see the data in ParaView on the laptop that has been provided--the screen is gray and stays gray except for maybe a frame or two when I zoom in and out. I plan on trying it out on my personal computer at home to see if I can actually view the data in the viewing plane.

Oh, and please forgive me if I used the dash incorrectly. We just covered dashes in my English class and I was debating over whether to use a semi-colon, a colon, a dash, or parenthesis for that part.

If my timing is correct, then I believe that I only have two more weeks at the OSC left.

I hope that everyone has a good weekend and I look forward to next week’s challenges!

Sincerely,
Austin Kochs
Metro High School Student Intern