Optimizing I/O Performance of HPC Applications with Autotuning
- Univ. of Illinois at Urbana-Champaign, IL (United States)
- Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
- Univ. of Illinois at Urbana-Champaign, IL (United States); Argonne National Lab. (ANL), Argonne, IL (United States)
Parallel Input output is an essential component of modern high-performance computing (HPC). Obtaining good I/O performance for a broad range of applications on diverse HPC platforms is a major challenge, in part, because of complex inter dependencies between I/O middleware and hardware. The parallel file system and I/O middleware layers all offer optimization parameters that can, in theory, result in better I/O performance. Unfortunately, the right combination of parameters is highly dependent on the application, HPC platform, problem size, and concurrency. Scientific application developers do not have the time or expertise to take on the substantial burden of identifying good parameters for each problem configuration. They resort to using system defaults, a choice that frequently results in poor I/O performance. We expect this problem to be compounded on exascale-class machines, which will likely have a deeper software stack with hierarchically arranged hardware resources.We present as a solution to this problem an autotuning system for optimizing I/O performance, I/O performance modeling, I/O tuning, and I/O patterns. We demonstrate the value of this framework across several HPC platforms and applications at scale.
- Research Organization:
- Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
- Sponsoring Organization:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)
- DOE Contract Number:
- AC02-05CH11231
- OSTI ID:
- 1825486
- Journal Information:
- ACM Transactions on Parallel Computing, Vol. 5, Issue 4; ISSN 2329-4949
- Publisher:
- Association for Computing Machinery
- Country of Publication:
- United States
- Language:
- English
Minerva: An automated resource provisioning tool for large-scale storage systems
|
journal | November 2001 |
Taming parallel I/O complexity with auto-tuning
|
conference | November 2013 |
I/O acceleration with pattern detection
|
conference | January 2013 |
Automatic parallel I/O performance optimization in Panda
|
conference | January 1998 |
Improving parallel I/O autotuning with performance modeling
|
conference | January 2014 |
Skel: Generative Software for Producing Skeletal I/O Applications
|
conference | December 2011 |
Optimizing matrix multiply using PHiPAC: a portable, high-performance, ANSI C coding methodology
|
conference | January 1997 |
Performance modeling for the panda array I/O library
|
conference | January 1996 |
Breaking the Cloud Parameterization Deadlock
|
journal | November 2003 |
VORPAL: a versatile plasma simulation code
|
journal | May 2004 |
Omnisc'IO: A Grammar-Based Approach to Spatial and Temporal I/O Patterns Prediction
|
conference | November 2014 |
Ultrahigh performance three-dimensional electromagnetic relativistic kinetic plasma simulation
|
journal | May 2008 |
A Comparison of Logical and Physical Parallel I/o pAtterns
|
journal | September 1998 |
Optimization of sparse matrix-vector multiplication on emerging multicore platforms
|
conference | January 2007 |
Improved parallel I/O via a two-phase run-time access strategy
|
journal | December 1993 |
A multi-level approach for understanding I/O activity in HPC applications
|
conference | September 2013 |
Online Adaptive Code Generation and Tuning
|
conference | May 2011 |
Modeling and Predicting Disk I/O Time of HPC Applications
|
conference | June 2010 |
Cost-intelligent application-specific data layout optimization for parallel file systems
|
journal | February 2012 |
PERI - auto-tuning memory-intensive kernels for multicore
|
journal | July 2008 |
An Overview of Evolutionary Algorithms for Parameter Optimization
|
journal | March 1993 |
I/O performance challenges at leadership scale
|
conference | January 2009 |
Lessons from characterizing the input/output behavior of parallel scientific applications
|
journal | June 1998 |
Similar Records
Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)
Institute for Sustained Performance, Energy, and Resilience (SuPER)