Efficient Run-Time System Support for High Performance Reliable Reconfigurable Systems
This paper presents and demonstrates efficient run-time system support for high performance reliable reconfigurable systems. Central to this is a novel inter-task communication mechanism which uses the FPGA's configuration port as an inter-task communication medium instead of conventional
routing resources on the logic fabric. As a result, complex inter-task routing is completely eliminated, and tasks can be easily relocated across a chip in order to maximize performance and/or avoid emerging faults on-chip. This architecture allows for scalable, multi-task and multi-user applications
on FPGAs, whereby hardware tasks behave like software tasks. The paper demonstrates this in the context of a real hardware implementation of a widely used data mining application, namely K-Nearest Neighbor Classifier (KNN) which achieves considerably high performance and reliability
at a low speed overhead and footprint.
Keywords: DATA MINING; FAULT-TOLERANCE; FPGA; RECONFIGURABLE COMPUTING; RECONFIGURABLE SYSTEM; SYSTEM ON CHIP
Document Type: Research Article
Publication date: 01 December 2012
- JOURNAL OF COMPUTATIONAL INTELLIGENCE AND ELECTRONIC SYSTEMS publishes emerging research in the areas of the computational intelligence and electronic systems. JCIES publishes in all areas of computational intelligence design and applications: applications oriented developments, successful industrial implementations, design tools, technology reviews, computational intelligence education, and applied research, specific emphasis on power electronics, embedded systems, semiconductor devices, analogue circuits, digital electronics, microwave and millimeter-wave techniques, wireless and optical communications, sensors, instrumentation and medical electronics much more.
- Editorial Board
- Information for Authors
- Subscribe to this Title
- Aims and Scope
- Ingenta Connect is not responsible for the content or availability of external websites
- Access Key
- Free content
- Partial Free content
- New content
- Open access content
- Partial Open access content
- Subscribed content
- Partial Subscribed content
- Free trial content