Hot spot analysis in large scale shared memory multiprocessors
- Univ. of Toronto, Ontario (Canada). Computer Systems Research Institute
Scalable multiprocessors that support a shared-memory image to application programmers are typically based on physical memory modules that are distributed. Consequently, the access times for a particular processor to various parts of physical memory differ. In this paper, the authors explore the implications of this non-uniformity in memory access times. In particular, they study the effect of hot-spots in hierarchical large scale NUMA multiprocessors. Hot-spot analysis is of interest because coordinated threads of parallel programs lead to hot spots whose impact on performance may be substantial or even dominant. They have developed an analytical model of access latencies and contention for shared resources in the interconnection network that links the processors and memory modules. Their objective is to provide a better understanding of non-uniform memory access times in scalable architectures. They show the extent to which a variable can be shared before it becomes a performance bottleneck, and assess the potential gain from replication of shared data items. They also demonstrate that the backoff value (after a memory request rejection) must be chosen carefully to balance memory access time and network utilization. Finally, they show that memory utilization is improved by allowing memory request buffering.
- OSTI ID:
- 46284
- Report Number(s):
- CONF-931115--
- Country of Publication:
- United States
- Language:
- English
Similar Records
The performance implications of thread management alternatives for shared-memory multiprocessors
Distributing hot-spot addressing in large-scale multiprocessors