Parallel-aware, dedicated job co-scheduling within/across symmetric multiprocessing nodes
- Livermore, CA
- Kingston, NY
- Austin, TX
- Saugerties, NY
In a parallel computing environment comprising a network of SMP nodes each having at least one processor, a parallel-aware co-scheduling method and system for improving the performance and scalability of a dedicated parallel job having synchronizing collective operations. The method and system uses a global co-scheduler and an operating system kernel dispatcher adapted to coordinate interfering system and daemon activities on a node and across nodes to promote intra-node and inter-node overlap of said interfering system and daemon activities as well as intra-node and inter-node overlap of said synchronizing collective operations. In this manner, the impact of random short-lived interruptions, such as timer-decrement processing and periodic daemon activity, on synchronizing collective operations is minimized on large processor-count SPMD bulk-synchronous programming styles.
- Research Organization:
- Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- W-7405-ENG-48
- Assignee:
- Lawrence Livermore National Security, LLC (Livermore, CA)
- Patent Number(s):
- 7,810,093
- Application Number:
- 10/989,704
- OSTI ID:
- 1016148
- Country of Publication:
- United States
- Language:
- English
Effective distributed scheduling of parallel workloads
|
conference | January 1996 |
Fast collective operations using shared and remote memory access protocols on clusters
|
conference | January 2003 |
Operating system support for parallel programming on RP3
|
journal | September 1991 |
Dynamic coscheduling on workstation clusters
|
book | January 1998 |
Similar Records
Reducing communication in algebraic multigrid with multi-step node aware communication
Reducing communication in algebraic multigrid with multi-step node aware communication