skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Discrete Event Execution with One-Sided and Two-Sided GVT Algorithms on 216,000 Processor Cores

Journal Article · · ACM Transactions on Modeling and Computer Simulation
DOI:https://doi.org/10.1145/2611561· OSTI ID:1132340
 [1];  [1];  [1]
  1. ORNL
+ Show Author Affiliations

Global virtual time (GVT) computation is a key determinant of the efficiency and runtime dynamics of parallel discrete event simulations (PDES), especially on large-scale parallel platforms. Here, three execution modes of a generalized GVT computation algorithm are studied on high-performance parallel computing systems: (1) a synchronous GVT algorithm that affords ease of implementation, (2) an asynchronous GVT algorithm that is more complex to implement but can relieve blocking latencies, and (3) a variant of the asynchronous GVT algorithm to exploit one-sided communication in extant supercomputing platforms. Performance results are presented of implementations of these algorithms on up to 216,000 cores of a Cray XT5 system, exercised on a range of parameters: optimistic and conservative synchronization, fine- to medium-grained event computation, synthetic and non-synthetic applications, and different lookahead values. Performance of up to 54 billion events executed per second is registered. Detailed PDES-specific runtime metrics are presented to further the understanding of tightly-coupled discrete event dynamics on massively parallel platforms.

Research Organization:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States). Oak Ridge Leadership Computing Facility (OLCF)
Sponsoring Organization:
USDOE Office of Science (SC)
DOE Contract Number:
DE-AC05-00OR22725
OSTI ID:
1132340
Journal Information:
ACM Transactions on Modeling and Computer Simulation, Vol. 24, Issue 3; ISSN 1049-3301
Publisher:
Association for Computing Machinery
Country of Publication:
United States
Language:
English

References (14)

Scalable Time Warp on Blue Gene Supercomputers
  • Bauer Jr., David W.; Carothers, Christopher D.; Holder, Akintayo
  • 2009 ACM/IEEE/SCS 23rd Workshop on Principles of Advanced and Distributed Simulation (PADS) https://doi.org/10.1109/PADS.2009.21
conference June 2009
On deciding between conservative and optimistic approaches on massively parallel platforms conference December 2010
DSIM: Scaling Time Warp to 1,033 Processors conference January 2005
Reducing null messages in Misra's distributed discrete event simulation method journal January 1990
MPI Collective Communications on The Blue Gene/P Supercomputer: Algorithms and Optimizations
  • Faraj, Ahmad; Kumar, Sameer; Smith, Brian
  • 2009 17th Annual IEEE Symposium on High-Performance Interconnects (HOTI), 2009 17th IEEE Symposium on High Performance Interconnects https://doi.org/10.1109/HOTI.2009.12
conference August 2009
Computing global virtual time in shared-memory multiprocessors journal October 1997
Virtual time journal July 1985
Efficient Algorithms for Distributed Snapshots and Global Virtual Time Approximation journal August 1993
Middleware for real-time distributed simulations
  • McLean, Thom; Fujimoto, Richard; Fitzgibbons, Brad
  • Concurrency and Computation: Practice and Experience, Vol. 16, Issue 15 https://doi.org/10.1002/cpe.874
journal January 2004
The cost of conservative synchronization in parallel discrete event simulations journal April 1993
Scaling time warp-based discrete event execution to 104 processors on a Blue Gene supercomputer conference January 2007
GVT algorithms and discrete event dynamics on 129K+ processor cores conference December 2011
Discrete event modeling and massively parallel execution of epidemic outbreak phenomena journal July 2011
Providing determinism in the Time Warp operating system-costs, benefits, and implications conference January 1990

Similar Records

Related Subjects