[Halld-offline] GlueX at SC18

[Shepherd, Matthew]

Hi all,

I just wanted to share a photo showing that GlueX was featured as part of the Indiana U. exhibit at the Supercomputing 2018 conference that is ongoing now in Dallas.

The IU research technologies team was looking for a science application to showcase a 400 Gbps network link from IU to Chicago and then pieced together as 4 x 100 Gbps to the exhibit floor in Dallas where there is a rack of 32 x 32 nodes.  So we put 70 TB of REST data on the disk back home and are trying a simple analysis launch just to see how fast we can do it.  The use case that we are trying to demonstrate that there are certain types of jobs that are I/O bound and have large amounts of data where it would be nice to have one copy of the data in one place and have anyone with any cluster be able to compute against it from anywhere.  This is a larger scale demonstration of a system that is already in place on the IU campus that we use regularly to store data on a completely different system from our physics cluster but use the physics cluster for computing.  (When I left last night they were having some issues with the network link so they haven't yet run our test application.)

They also wanted some eye candy to go along with this (apparently people don't get as excited about histograms as we do), so I worked with Thomas to get an offline version of the 3D event display and cache of about 300 events.  We have a whole division in our research computing arm dedicated to visualization called the "Advanced Visualization Lab."  As you can see in the photo, they have complete disregard for gravity, but they where happy to get the event display running which made a nice conversation piece for people walking by.

They offered to help consult on any completing/polishing of the HDvis system -- some effort at the level of a couple of days is apparently easy to arrange without cost.  We have an in house expert in three.js had no trouble reading and making edits to the existing code.  He was going to try to implement gyro paddles to let people manipulate the detector by moving their hands, but didn't quite have the time.  (I'll be in touch with Thomas more about this.  Maybe he wants to make a visit.  Bloomington is beautiful in the winter.)

They have also offered to help make a "matterport" scan of Hall D and the tagger area.  For an idea of what this looks like, you can see this example:

https://my.matterport.com/show/?model=cRuWEHi2NwT

This could be incredibly useful for showing people the hall when the hall is not accessible.  (Or when you want to show the hall to some new grad student in your office hundreds of miles away from JLab.)  Apparently it also supports stereo so you can do a VR walkthrough of the hall with your phone and a stereo viewer.  It takes some time to do the photography so we'll have to coordinate schedules with hall accessibility, but I really think this would be a good time investment at some point.  The local IU expert seemed really excited to try to do it.

Matt



---------------------------------------------------------------------
Matthew Shepherd, Professor
Department of Physics, Indiana University, Swain West 265
727 East Third Street, Bloomington, IN 47405

Office Phone:  +1 812 856 5808

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20181115/317f29b1/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: IMG_2889.jpeg
Type: image/jpeg
Size: 221539 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20181115/317f29b1/attachment-0001.jpeg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4019 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20181115/317f29b1/attachment-0001.p7s>