S
Sameer Shende
Researcher at University of Oregon
Publications - 125
Citations - 4661
Sameer Shende is an academic researcher from University of Oregon. The author has contributed to research in topics: Instrumentation (computer programming) & Profiling (computer programming). The author has an hindex of 26, co-authored 121 publications receiving 4384 citations. Previous affiliations of Sameer Shende include Forschungszentrum Jülich & Los Alamos National Laboratory.
Papers
More filters
Book ChapterDOI
Workload characterization using the TAU performance system
TL;DR: This paper demonstrates workload characterization features in the TAU parallel performance system for elucidating the performance of the MPI library based on the sizes of messages and discusses the use of mapping for memory consumption characterization.
Book ChapterDOI
Improving the scalability of performance evaluation tools
TL;DR: The extensions to TAU to contend with large data volumes associated with increasing core counts are described, including new instrumentation choices, efficient handling of disk I/O operations in the measurement layer, and strategies for visualization of performance data at scale in TAU's analysis layer.
Journal ArticleDOI
Checkpoint/restart approaches for a thread-based MPI runtime
Julien Adam,Maxime Kermarquer,Jean-Baptiste Besnard,Leonardo Bautista-Gomez,Marc Pérache,Patrick Carribault,Julien Jaeger,Allen D. Malony,Sameer Shende +8 more
TL;DR: The work done in the MPI runtime to enable both transparent and application-level checkpointing mechanisms is described, and it is shown how existing checkpointing methods can be practically applied to a thread-based MPI implementation given sufficient runtime collaboration.
Performance Tools for Parallel Java Environments
Sameer Shende,Allen D. Malony +1 more
TL;DR: The experience in extending the TAU performance system to a parallel Java environment based on mpiJava is related, the instrumentation model used, how performance measurements are made, and the overhead incurred are described.
Book ChapterDOI
Early experiences with KTAU on the IBM BG/L
TL;DR: KTAU extends the TAU performance system with kernel-level monitoring, while leveraging TAU's measurement and analysis capabilities, and reports early experiences using KTAU in ZeptoOS on the IBM BG/L system.