NPPS Leadership Team

[Brett Viren <bv AT bnl.gov>]

Here's a minor FYI but comments are welcome.

I am generally supportive of the proposal.  I commented on the google
doc and replied to the email thread that Kirby may wish to include
adding HDF5 support to Wire-Cell Toolkit (for support of Machine
Learning training on HPC) as part of his project #1.  This is something
I intend to provide myself but would be happy if someone else did it.
Kirby was receptive to the idea.

Meifeng pointed out that proposed project #2 is complimentary to the CCE
work CSI is already doing on evaluating Wire-Cell Toolkit for porting to
HPC/GPU.

For context, those named are "benedikt.hegner AT cern.ch"
<benedikt.hegner AT cern.ch>, "e.lancon AT gmail.com" <e.lancon AT gmail.com>,
"kleese AT bnl.gov" <kleese AT bnl.gov>, Meifeng Lin <mlin AT bnl.gov>,
"abidmuslim AT gmail.com" <abidmuslim AT gmail.com> and cc'ed Erica Snider
<erica AT fnal.gov> (LArSoft lead).

-Brett.


Michael H Kirby <kirby AT fnal.gov> writes:

> Hi Abid, Benedikt, Brett, Eric, Kerstin, and Meifing, 
>
> I wanted to send a draft of the document that Erica Snider and I put 
> together at the request of Liz
> Sexton-Kennedy and Jim Amundson. I'd attended some of the CCE coordination 
> meetings and taken the list
> of people on this email from the basecamp website for those meetings. The 
> document tries to build off the
> proposal of the Energy Frontier document, and propose two projects for 
> Intensity Frontier:
>
> - event data model development
> - accelerator-optimized algorithms through high-level, parallelization 
> libraries
>
> I wanted to have some input from people at Brookhaven because we included 
> 2D deconvolution of LArTPC
> signals as a use case for Project 2, and specifically called at WireCell 
> Toolkit as a specific implementation of
> a 2D deconvolution+data prep algorithm. I'm hoping that you would support 
> having WireCell as something
> to well matched for CCE funding and support for application of those 
> parallelization libraries. If so, we're
> hoping to get these use cases included in slides presented at DoE tomorrow. 
> The link below is open for
> comments and edits.
>
> https://docs.google.com/document/d/173mRSL0VMXyAaI3HCD0yc-ttXl9whmZeEB_HSpLi7h0/edit?usp=sharing
>
>
> Cheers,
> Kirby
>
> Michael Kirby
> Scientist I
>
> Scientific Computing Division
> Fermi National Accelerator Laboratory
> PO Box 500, MS 120, Batavia, IL 60510, USA
> kirby AT fnal.gov
> Office: 630.840.4326
> Cell: 630.965.1456
>
> Connect with us!
> Newsletter | Facebook | Twitter

Attachment: signature.asc
Description: PGP signature