Abstract
In a recent paper. Pfister and Norton[4] brought to our attention the potential degradation in performance of multi-stage interconnection networks in the presence of nonuniform traffic patterns. More specifically, they studied the effect of a single hot spot - a location that is accessed more frequently - in an otherwise uniform traffic environment. Their simulations assumed that all input lines contribute to the hot spot. Their results showed that the effect of hot spots drastically reduced the bandwidth of the network. In this note. we confirm Pfister and Norton's results and present supplemental data and arguments which illustrate other aspects of hot spot contention. In particular, we show that the performance improves when the source of the hot spot is more localized.
Index Terms
- On “hot spot” contention
Recommendations
Comparative Performance Evaluation of Hot Spot Contention Between MIN-Based and Ring-Based Shared-Memory Architectures
Hot spot contention on a network-based shared-memory architecture occurs when a large number of processors try to access a globally shared variable across the network. While Multistage Interconnection Network (MIN) and Hierarchical Ring (HR) structures ...
Code Hot Spot: A tool for extraction and analysis of code change history
ICSM '11: Proceedings of the 2011 27th IEEE International Conference on Software MaintenanceCommercial software development teams have limited time available to focus on improvements to their software. These teams need a way to quickly identify areas of the source code that would benefit from improvement, as well as quantifiable data to defend ...
Fuzzy Hot Spot Identification for Big Data: An Initial Approach
2019 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)Hot spot identification problems are present across a wide range of areas, such as transportation, health care and energy. Hot spots are locations where a certain type of event occurs with high frequency. A recent big data approach is capable of ...
Comments