Final Program

Conference Location: Ohio Supercomputer Center's Bale Theater

Monday, August 26th

07:45 – Registration
7:45 – 8:20 Continental Breakfast
8:20 – 8:30 Opening Remarks
8:30 – 9:30
Keynote Talk
Experiences Using MVAPICH in a Production HPC Environment at TACC
Karl Schulz, Texas Advanced Computing Center
9:30 – 10:00
Building Multi-Petaflop Systems with MVAPICH2 and Global Arrays
Abhinav Vishnu, Pacific Northwest National Laboratory
10:00 – 10:30
MVAPICH Saves the World!
Adam Moody, Lawrence Livermore National Laboratory
10:30 – 11:00 Break
11:00 – 11:30
Overview of MVAPICH2 and MVAPICH2-X: Latest Status and Future Roadmap
Dhabaleswar K. (DK) Panda, The Ohio State University
11:30 – 12:00
Optmization and Tuning of Job Start-up, Process Mapping, and Point-to-Point Communication in MVAPICH2
Hari Subramoni & Sreeram Potluri, The Ohio State University
12:00 – 1:00 Lunch
1:00 – 1:30
MVAPICH2 with Dual Rail 3D Torus Support on SDSC Data Intensive Gordon Cluster
Mahidhar Tatineni and Amit Majumdar, San Diego Supercomputer Center
1:30 – 2:00
Optimization and Tuning of Hybrid mode, Multirail, 3D Torus Support and QoS in MVAPICH2
Hari Subramoni, The Ohio State University
2:00 – 2:30
Experience with MVAPICH2 on Mellanox Hardware
Gilad Shainer, Mellanox Technologies
2:30 – 3:00
GPU Aware MPI and Applications: Stencil Computations
Bianco Mauro, Swiss National Supercomputing Center, Switzerland
3:00 – 3:30 Break
3:30 – 4:10
Contributed Presentations (Session I)

What Communication Library Can do with a Little Hint from Programmers?
Takeshi Nanri, Kyushu University and JST CREST, Japan

Benefits of Kernel-assisted Approaches on Scientific Applications: An Analysis Using LIMIC2 and CMA in MVAPICH2
Jerome Vienne, Texas Advanced Computing Center
4:10 – 4:40
Troubleshooting Guidelines for Installing and Using MVAPICH2 and MVAPICH2-X
Devendar Bureddy, The Ohio State University
4:40 – 5:20 Open Mic Session
6:30 – Reception and Dinner at Bravo Restaurant
1803 Olentangy River Rd
Columbus, OH 43212

Tuesday, August 27th

07:45 – Registration
7:45 – 8:30 Continental Breakfast
8:30 – 9:15
Optimization and Tuning of Collectives in MVAPICH2
Hari Subramoni, The Ohio State University
9:15 – 10:00
Optimizing and Tuning of GPU and Xeon Phi Support in MVAPICH2
Sreeram Potluri, The Ohio State University
10:00 – 10:30 Break
10:30 – 11:20
Contributed Presentations (Session II)

Performance Analysis of Scientific Applications on GPUs using MVAPICH2
Mahidhar Tatineni and Dongju Choi, San Diego Supercomputing Center

High-Performance Vectorized Discrete Approximation on GPU Clusters by MVAPICH2
Di Zhao, The Ohio State University
11:20 – 11:50
Using MVAPICH2-X for Hybrid MPI + PGAS (OpenSHMEM and UPC) Programming
Jithin Jose, The Ohio State University
11:50 – 12:20
Fault-Tolerance Support in MVAPICH2
Raghu Rajachandrasekar, The Ohio State University
12:20 – 12:35
Discussion on Future MUG Events
12:35 – 1:30 Lunch
1:30 – 3:30
Interactive/Hands-on Session with MVAPICH2 and MVAPICH2-X Developers