Talks will take place the Center for Computational Sciences, Boston University, 3 Cummington Mall, PRB (Physics Research Building) Room 595. The plan is to have short presentations to raise questions and provide ample time for discussion by all participants.
Remote Access
Audio: Call into +1-866-740-1260, access code 3640117.
Slides: Go to www.readytalk.com, access code 3640117.
Tentative Schedule of Speakers & Panelists
Nov. 4 , 2013
8:30 AM Coffee et al
9:00 AM Start of Morning Session
- Anshu Dubey (LBL) Summary of first two SIMAC workshops
- Mike Clark (NVIDIA) Lattice QCD algorithms at the Exascale
- Joe Eaton (NVIDIA) NVIDIA MPI-enabled Iterative Solvers for Large Scale Problems
- Frank Winter (Jefferson Laboratory): Accelerating Lattice QCD Calculations on GPU-Enabled Systems
Open Discussion:
- How do we accelerate the transition to Heterogeneous Architectures with Accelerators?
12:30 PM Lunch PRB365
1:30 PM Start of Afternoon Session
- Andrew Pochinsky (MIT) Qlua: Scripting a DSL
- James Brannick (Penn State) AMG at the Exascale?
- J. (Ram) Ramanujam (LSU) Software for High Level Optimizing of PDE.
Open Discusion:
- What can languages do to help? (lead by Shoaib Kamil)
- Array Construction in various Languages: Matrices Vs Meshes (summary Anshu)
- AMR implementations in different packages (summary Petros)
5:00 PM Panel Discussion: Professional Development and maximizing the impact of a SIMAC Institute.
Hans Johansen (LBL) Chair
Lori Diachin (LNLL),
Bruce Boghosian (Tufts)
Roscoe Giles(BU),
Claudio Rebbi(BU)
6:30 PM Reception and Buffet Dinner at Hariri Institute for Computing and Computational Science & Engineering
111 Cummington Mall, Room 180. See Travel & Hotel page for an updated map including the Hariri Institute, or click here for an alternative map for the dinner specifically.
Nov 5 , 2013
9:00 AM Start of Morning Session
- Chris Hill (MIT) Multi-scale Fluids and co-processor Hardware
- Lorena Barba (George Washington) Fast Multipole Methods in the Exascale Era
- Erik Schnetter (Perimeter Institute) Adaptive Mesh Refinement: Current state in Cactus, and future possibilities
- Bryce Lelbach & Dominic Marcello (LSU) Message-driven AMR Techniques for Binary Star Simulations.
12:30 Round Table Discussion : How Can the Industry Engage with and Benefit from a Software Institute Like SIMAC
Don Lamb (U of Chicago) Chair
Andrew Jones (NAG),
Jonathan Cohen (NVIDIA),
Josh Danczyk (Caterpillar)
James F Wiedenhoefer (GE Global Research) et al
1:00 Lunch PRB261