[Openchemistry-developers] Fwd: GSoC '16 Proposal

[Anish Pednekar]

---------- Forwarded message ----------
From: *Anish Pednekar* <pednekar.anish at gmail.com>
Date: Thursday 24 March 2016
Subject: GSoC '16 Proposal
To: Geoffrey Hutchison <geoff.hutchison at gmail.com>


Thank you for your response. I have gone through the links and I would like
to ask if we are expected to do both i.e adapting OpenMM for other
molecular fields as well as writing a new implementation for the existing
OpenBabel code to speed up the calculations using OpenMP/OpenCL?

On Wed, Mar 23, 2016 at 9:31 PM, Geoffrey Hutchison <
geoff.hutchison at gmail.com
<javascript:_e(%7B%7D,'cvml','geoff.hutchison at gmail.com');>> wrote:

> > Hi, I am Anish Pednekar, a second year student studying Information
> Technology at Goa College of Engineering, India. I'd like to contribute to
> openchemistry through Google Summer of Code. I have gone through the GSoC
> ideas page and the project - GPU and Multi-Core Enabled High Performance
> Force Field Calculations interested me.
> >
> > I have worked with Cuda and OpenCL before and have successfully
> completed courses on Udacity and Coursera. I served as a community TA for
> the Heterogenous Parallel Programming course on Coursera. I am also
> familiar with Git.
>
> We have a guide on applying here:
> http://wiki.openchemistry.org/Applying_to_GSoC
>
> There are several excellent guides to molecular mechanics force fields out
> on the web, but you might want to look here:
> http://vergil.chemistry.gatech.edu/courses/chem6485/pdf/molmech-lecture.pdf
>
> There are two approaches to consider for the project:
> 1) Adapting an existing force field package (e.g., OpenMM) for other
> molecular force fields - for example using Open Babel's current force field
> implementation to export parameters and custom force types to OpenMM XML
> format.
>
> 2) Replacing the current code in Open Babel with a new implementation that
> is designed to be optimization-friendly, vectorized, and OpenMP-ready.
> (i.e., it breaks the molecule into work-groups to run on OpenMP or OpenCL
> threads).
>
> Hope that helps,
> -Geoff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/openchemistry-developers/attachments/20160325/7ed96668/attachment.html>