You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

Present: Henry van den Bedem, Joe Blaylock, Arno Candel, Igor Gaponenko, Vineet Rawat, Bebo White

At the start of the meeting, Arno presented the following collection of candidates:

*****************************************************************************************************************************************************************************************************

Candidates as of March 7 2011:

Horst Simon suggests:

* himself (Deputy Lab director – LBNL, Director NERSC) – Status of HPC in the US

* Peter Nugent (Group leader - LBNL) – Cosmology (OK from Ralf Kaehler and Tom Abel)

* Julian Borrill (Staff scientist - LBNL) – Cosmology (OK from Ralf Kaehler and Tom Abel)

* Jamie Sethian (Prof. - UC Berkeley) – General Applied Math and Sci Comp

* Filipe Maia (Postdoc - LBNL) – GPUs/Hawk (Igor: might be too technical)

* Jim Demmel (Prof. - UC Berkeley) - Optimizing Performance by Avoiding Communication

* Andrew Canning (Sen. Staff - LBNL) - Mat Science on Many Core Architectures (PARATEC)

* Sam Williams (PhD08 - LBNL) - Auto-Tuning Memory-Intensive Kernels for Multicore 

Jakec Becla suggests:

* Jeff Hammerbacher (wall street, facebook, cloudera,

   see [http://www.accel.com/bio/jeffhammerbacher.php|http://www.accel.com/bio/jeffhammerbacher.php])

* Oliver Ratzesberger (director of eBay analytics) 

Travis Brooks suggests:

* Facebook Data Science group (friends visualization etc)

* Eugene Kim (Wikimedia strategic planning) 

Igor Gaponenko suggests:

* David Patterson (CS Prof. - UC Berkeley)

Co-founder of ParLab @ UC Berkeley, Led the design of RISC 1 -> SPARC, Leader of RAID project. National Academy of Engineering, National Academy of Sciences, etc. 

Henry van den Bedem suggests:

* Tony DeRose from Pixar ([http://graphics.pixar.com/people/derose/index.html|http://graphics.pixar.com/people/derose/index.html]).

* Mohammed AlQuraishi  (Stanford) “De novo inference of protein-DNA energy potentials using compressed sensing"

Bebo White suggests:

* Danese Cooper – Wikimedia CTO

* Leo Paporte (leoville.com) – Online Social Media

* Jeff Johnson – Book “Designing with the Mind in Mind” (About Interfaces)

****************************************************************************************************************************************************************************************************************

During the meeting, a few new recommendations were made:

Henry suggested:

* Mike Huffer (SLAC, RCE for LSST/ATLAS)

* Todd Martinez (SLAC, GPUs for chemistry)

* SUNCAT (SLAC, catalysts for clean energy created using computational methods)

Igor suggested:

* Anton Barty (DESY Germany, work on LCLS imaging): 

[https://slacportal.slac.stanford.edu/sites/ad_public/events/FLS2010/Lists/WorkingGroup4/Attachments/23/Anton_Optics.pdf|https://slacportal.slac.stanford.edu/sites/ad_public/events/FLS2010/Lists/WorkingGroup4/Attachments/23/Anton_Optics.pdf]

[http://www.nature.com/nphoton/journal/v2/n7/pdf/nphoton.2008.128.pdf|http://www.nature.com/nphoton/journal/v2/n7/pdf/nphoton.2008.128.pdf]

MEETING OUTCOME:

1) We agreed that while the content should stay mostly in line with SLAC's research programs, a healthy mix of more general (but still strongly computing related) talks from outside speakers would greatly improve the changes for overall success.  Ratio 2:1 or so.

2) Frequency: Monthly OK, gives us freedom to sprinkle in the occasional surprise speaker

3) Time and location does not have to be strictly fixed, since it's monthly, people might forget / don't care.  We prefer the Kavli auditorium, if available.

4) Which day?

Monday: ruled out because of weekly colloquium at 4PM

Tuesday: OK?

Wednesday: ruled out because of weekly LCLS seminar (around 3PM)

Thursday: 3PM OK?

Friday: ruled out because of proximity to weekend

Proposed time line for the first seminars:

  • End of April (or as soon as Amber is here):  1st seminar: LCLS: Ideally, Anton Barty on imaging (suggested by Igor, seems ideal to us)
  • Mid of May: 2nd seminar: Cosmology: Peter Nugent or Julian Borrill from LBNL (suggested by Horst Simon, OK from Kavli colleagues)
  • End of May: sci comp workshop
  • Mid of June: 3rd seminar: General talk by someone in the movie/gaming industry (Pixar, NVidia or Electronics Arts)
  • Mid of July: 4th seminar: Mike Huffer (LSST/ATLAS), Todd Martinez (GPU chemistry), David Patterson (Comp Science Prof. LBNL), FaceBook Data Science would all be suitable candidates. But might skip because too close to holidays (or schedule a bit earlier?)

Action items:

* Henry to check with Pixar

* Igor to check with Anton Barty - since from Germany collaborating with LCLS: it's not clear that he will be here at the end of April/Early May, then we might use him as a "Joker" in between regular talks.

* Arno: As soon as possible, contact Cosmology speakers from LBNL and identify the first speaker if Anton Barty is not available.

  • No labels