Reducing the variance of point-to-point transfers for parallel real-time programs
- 1 January 1994
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Parallel & Distributed Technology: Systems & Applications
- Vol. 2 (4) , 20-31
- https://doi.org/10.1109/88.345963
Abstract
Investigations that analyze the time an operating system takes to schedule, interrupt and "context-switch" to another process or job have helped developers produce highly optimized and tuned operating systems that can provide more than 99% sustained processor use for most uniprocessor applications. However, when these operating systems are installed on CPUs that are interconnected with a low-latency (user-space) communication mechanism, large variances typically occur in the time it takes to send a point-to-point message. In this article, we examine how to reduce the difference between worst-case and average-case message latency that can contribute to variance in fine-grain parallel programs. Changing how the operating system handles interrupt processing and scheduling can greatly reduce the difference between these latencies, thus increasing a program's performance.Keywords
This publication has 5 references indexed in Scilit:
- The Shark continuous-media file serverPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Architecture and implementation of VulcanPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Reducing the variance of point to point transfers in the IBM 9076 parallel computerPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- A real-time locking protocolIEEE Transactions on Computers, 1991
- Aperiodic task scheduling for Hard-Real-Time systemsReal-Time Systems, 1989