BLIS Retreat
Preliminary Program
Thursday Sept. 5
| Morning | POB 2.402 | ||
| 8:30-9:00 | Breakfast (Outside POB 2.402) | ||
| 9:00-9:05 | Welcome | Robert van de Geijn | |
| 9:05-10:00 | BLIS overview | SLIDES | Field Van Zee |
| 10:00-10:20 | Parallelizing BLIS | SLIDES | Tyler Smith |
| 10:20-11:00 | Break (POB 2.402) | ||
| 11:00-11:30 | Developing low-level assembly kernels with Peach-Py | SLIDES | Marat Dukhan |
| 11:30-11:50 | BLIS as a Research Vehicle | SLIDES | Bryan Marker |
| 11:50-12:10 | Porting BLIS to new architectures. Early experiences | SLIDES | Francisco Igual |
Lunch (GDC 6.302 - Computer Science Faculty Lounge)
| Afternoon | Various rooms GDC fifth floor, south building | ||
| 1:30-4:30 | Hands-on experience with BLIS on your favorite
architecture (you will have a chance to
work with those who have tried it, in small groups.) Materials Octave script example | ||
| 3:00 | Break (5th floor south, kitchen area) | ||
| 4:30-5:00 | BLIS: Moving forward (Robert van de Geijn) | ||
| 5:00- | John suggests we visit Gabriel's Cafe at the AT&T Center (Sorry, can't charge that to a grant...) | ||
Friday Sept. 6
| Morning | POB 2.402 | ||
| 8:30-9:00 | Breakfast (Outside POB 2.402) | ||
| 9:00-9:30 | Effective Methods for Propagating the Schroedinger Equation in Time | SLIDES | Barry Schneider |
| 9:30-10:00 | Distributed Contraction of Symmetric Tensors | Saday Sadayappan | |
| 10:00-10:20 | When all you have is linear algebra, everything looks like a matrix | SLIDES | Devin Matthews |
| 10:20-11:00 | Break (POB 2.402) | ||
| 11:00-11:20 | Optimizing the inner kernel: a low power, high performance core for matrix computations | SLIDES | Ardavan Pedram |
| 11:20-11:50 | Performance Modeling for DLA Kernels | SLIDES | Elmar Peise |
| 11:50-12:10 | Discussion of requirements for the BLIS Fortran interface(s) | MATERIALS | Jeff Hammond |
Lunch (GDC 6.302 - Computer Science Faculty Lounge)
| Afternoon | POB 2.402 | ||
| 1:30-1:50 | Compositional Parallelism: How? | Tim Mattson | |
| 1:50-2:10 | Lessons from a parallel sparse direct solver with multilevel scheduling | SLIDES | Kyungjoo Kim |
| 2:10-2:40 | Performance-portable kernels in OpenCL: Lessons learned | Karl Rupp | |
| 2:40-3:10 | CANCELLED | Greg Henry | |
| 3:10-4:30 | Discussion/Coffee (POB 2.402) | ||
| 4:30 | Wrapup | ||