BLIS Retreat 2016
Program (First Draft)
Contributed talksMonday Sept. 19
| Morning | POB 2.402 | ||
| 8:30-9:00 | Breakfast (POB 2.402) | ||
| 9:00-9:30 | BLIS: Year In Review, 2015-2016 | SLIDES | Field Van Zee, UT-Austin |
| 9:30-10:10 | Implementing Strassen-like Fast Matrix Multiplication Algorithms with BLIS | SLIDES | Jianyu Huang and Leslie Rice, UT-Austin |
| 10:10-10:40 | A New I/O Lower Bound for GEMM with a Tight Constant | For SLIDES contact speaker | Tyler Smith, UT-Austin |
| 10:30-11:00 | Coffee (POB 2.402) | ||
| 11:00-11:30 | Scalable Dense Matrix Multiplication on Multi-Socket Many-Core Systems with Fast Shared Memory | SLIDES | Natalia Vassilieva, Hewlett Packard Labs |
| 11:30-12:00 | An Implementation of GEMM for DMA-enabled Architectures | SLIDES | Devangi Parikh, TI |
| 12:00-12:30 | BLAS for Deep Learning: tuple, mixed-precision, fixed-point, and binary GEMM | SLIDES | Marat Dukhan, GATech |
12:30 - 2 Lunch (GDC 6.302 - Computer Science Faculty Lounge)
| Afternoon | POB 2.402 | ||
| 2:00-2:30 | A Case Study for Malleable Thread-Level Linear Algebra Libraries: The LU Factorization | SLIDES | Sandra Catalan, Univ. Jaume I |
| 2:30-3:00 | libFLAME Optimizations with BLIS | SLIDES | Kiran Varaganti, AMD |
| 3:00-3:30 | Extended BLAS, Integer BLAS, Batched BLAS, etc. | Greg Henry, Intel | |
| 3:30-4:00 | Coffee (POB 2.402) | ||
| 4:00-4:30 | An Algorithmic Specific Code Generator for Matrix-Matrix Multiply-Like Operations | SLIDES | Richard Veras, CMU |
| 4:30-5:00 | A BLIS affair with FPGAs |
For VIDEO contact speaker SLIDES | Tze Meng Low |
| 5:00-5:30 | Cl1ck + LGen: FLAME for small scale linear algebra | SLIDES | Diego Fabregat, RWTH-Aachen University |
Tuesday Sept. 20
| Morning | POB 2.402 | ||
| 8:30-9:00 | Breakfast (POB 2.402) | ||
| 9:00-9:30 | Tensor Contraction with BLIS | SLIDES | Devin Matthews, UT |
| 9:30-10:00 | Design of a high-performance GEMM-like Tensor-Tensor Multiplication | SLIDES | Paul Springer, RWTH-Aachen University |
| 10:00-10:30 | Using BLIS for tensor computations in Q-Chem | SLIDES | Evgeny Epifanovsky, Q-Chem |
| 10:30-11:00 | Coffee (POB 2.402) | ||
| 11:00-11:30 | PeachPy.io: a platform for crowdsourcing performance tuning | SLIDES | Marat Dukhan, GATech |
| 11:30-12:00 | Dark Memory and Accelerator-Rich System Optimization in the Dark Silicon Era | SLIDES | Ardavan Pedram, Movidius and Stanford |
| 12:00-12:30 | A set of high performance kernel matrix operations on CPU, KNL and GPU | SLIDES | Chenhan Yu, UT-Austin |
| 12:30-12:45 | Closing comments | Robert van de Geijn, UT-Austin | |