Log in

Previous Entry | Next Entry

Second Annual NSF ITR KMC Workshop

Today we had the Second Annual Workshop on Learning-enhanced Adaptive Parallel Kinetic Monte Carlo (LEAP-KMC) simulation1, a collaborative National Science Foundation Information Technology Research (NSF ITR) project among KSU Physics, KSU CIS, the Physics Department at the University of Toledo, and the Faculty of Physics at the Russian Academy of Sciences.

Workshop Program
09:00-09:10 Opening remarks by Abdelkader Kara
09:10-09:30 Giridhar Nandipati, "Parallel KMC simulations of Cu(111) island coarsening using SLKMC database"
09:30-09:50 Yunsic Shim, "Parallel accelerated dynamics simulations of epitaxial growth"
09:50-10:10 Jacques G. Amar, "Strategies for efficient parallelization of KMC simulations"
10:10-10:30 Sarang Mairal, "Performance measurements in parallel computing"
10:30-10:40 Virgil Wallentine "Goals and perspectives for KMC parallel computing"

10:40-11:10 COFFEE BREAK

11:10-11:30 William Hsu, "Multiscale machine learning for LEAP-KMC energy estimation:experiments with genetic programming"
11:30-11:50 William Hsu and Andrew Walters, "Visualization of KMC simulation data and evolutionary computation: the LEAP infrastructure and content management system"
11:50-12:10 Altaf Karim, "Islands diffusion using SL-KMC"
12:10-12:30 Saad Ahmad, "Islands decay inside vacancy islands using SL-KMC"
12:30-12:50 Handan Yildirim, "Prefactors for diffusion mechanisms on Ag and Cu surfaces"


15:30-16:20 ITR and the Outreach program
    15:30-15:40 Nameer Baker, "Introduction to Nanomodling and the ROPE program"
    15:40-16:00 Justin Blair, Josh Lynch and Brad Carrender, "Modeling of a protein behaviour"
    16:00-16:20 Sara Peters, Xiao Li and Felix Wang, "Modeling of a crystal"
16:20-16:30 Concluding remarks by Talat S. Rahman

Yeah... this went a tad bit over (we actually finished right around 16:30, though, by deferring Altaf's talk and the ones after that to after lunch, and having those during the round table session from 14:15 on).

The KMC project addresses the problem of scaling up and speeding up atomic-scale simulations of nanomaterials processes such as formation of islands and thin films in metal vapor deposition. Deposition is used to manufacture semiconductor wafers and high-density optical recording media such as CD-Rs, DVD-Rs, and Blu=Ray. A long-term goal of the research is to extend this to other nanostructures such as C60 (also known as fullerenes or buckyballs) macromolecular structures in biochemistry: proteins and peptides, genomic structures (histones, chromatin, mRNA, etc.

The project started in January, 2004, when Talat, Kader, Virg, Jacques and I wrote a proposal to NSF. It was one of the last round of ITR grants to be awarded. Since then, we've been busy learning to speak one another's languages. On our en, this has necessitated our learning about solid state physics and simulation methods; on their end, parallel programming and algorithms, genetic and evolutionary computation, and machine learning.

Here are links to the nascent Tiki for the LEAP-KMC project/a>, built originally by Summer Undergraduate Research Opportunity Program (SUROP) participants Sadie L. Bell and Moises Figueroa, and to last year's workshop page. If you are interested in the parallel and distributed computing, machine learning, and GEC research and results that are coming out of our work, check this Tiki as we populate it with the slides from the workshop this week, and with preprints of papers over the summer.

The high school students did indeed work hard this year, and gave two rather nice talks. Kudos to Ahlam Al-Rawi, Talat's research associate, and her undergrad organizers, Asma (her daughter) and Nameer (Talat's son).

All in all, we're all pleased with the progress thus far, though there's a lot more to be done. martinsamuel, weninger, and I will be busy with this work in the coming year. I can tell you that I for one know how these kids felt:

Pictured, left to right: Xiao Li, Sara Peters and Felix Wang.
Click on the image for the full-size version.

1 There are several of us to credit or blame for that acronym. I believe it was Virg Wallentine, our department head, who suggested an acronym, Abdel Kader Kara, one of the Principal Investigators, who was calling it SL-KMC or ST-KMC for Self Learning/Teaching, and me who suggested "enhanced" because IIRC it was drifting towards Learning Adaptive Parallel and LeAP sounds better than LAP. Well, less unsavory, anyway.


Latest Month

December 2008

KSU Genetic and Evolutionary Computation (GEC) Lab



Science, Technology, Engineering, Math (STEM) Communities

Fresh Pages


Powered by LiveJournal.com
Designed by Naoto Kishi