Exploiting spatial parallelism in Ethernet-based cluster interconnects
Authors:
- 1 April 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 2008 IEEE International Symposium on Parallel and Distributed Processing
- p. 1-8
Abstract
In this work we examine the implications of building a single logical link out of multiple physical links. We use MultiEdge to examine the throughput-CPU utilization tradeoffs and examine how overheads and performance scale with the number and speed of links. We use low- level instrumentation to understand associated overheads, we experiment with setups between 1 and 8 1-GBit/s links, and we contrast our results with a single 10-GBit/s link. We find that: (a) Our base protocol achieves up-to 65% of the nominal aggregate throughput, (b) Replacing the interrupts with polling significantly impacts only the multiple link configurations, reaching 80% of nominal throughput, (c) The impact of copying on CPU overhead is significant, and removing copying results in up-to 66% improvement in maximum throughput, reaching almost 100% of the nominal throughput, (d) Scheduling packets over heterogeneous links requires simple but dynamic scheduling to account for different link speeds and varying load.Keywords
This publication has 12 references indexed in Scilit:
- MultiEdge: An Edge-based Communication Subsystem for Scalable Commodity Servers2007
- A scalable communication layer for multi-dimensional hyper crossbar network using multiple gigabit ethernet2006
- An Efficient Programmable 10 Gigabit Ethernet Network Interface Card2005
- Microbenchmark performance comparison of high-speed cluster interconnectsIEEE Micro, 2004
- Generalized inverse multiplexing of switched ATM connections2002
- The Quadrics network (QsNet): high-performance clustering technology2002
- The Quadrics network: high-performance clustering technologyIEEE Micro, 2002
- User-Space Communication: A Quantitative Study1998
- Fast messages: efficient, portable communication for workstation clusters and MPPsIEEE Concurrency, 1997
- Inverse multiplexingIEEE Communications Magazine, 1994