TY - JOUR
T1 - Synthesis of fault-tolerant and real-time microarchitectures
AU - Orailoǧlu, Alex
AU - Karri, Ramesh
N1 - Funding Information:
ACKNOWLEDGMENT This work was supported by the National Science Foundation Industry /University Cooperative Research Center on Ultra High Speed Integrated Circuits and Systems at the University of California, San Diego, California.
PY - 1994/4
Y1 - 1994/4
N2 - Microarchitectural implementations of real-time signal-processing algorithms for mission-critical applications (such as sensors on spacecrafts) are characterized by two unique requirements: real-time processing of a steady stream of input signals, which requires a high-performance implementation such as pipelining, and reliable operation over the mission lifetime, which mandates support for fault tolerance. We relate high-performance and fault-tolerance constraints to chip area and present a methodology for synthesizing area-efficient microarchitectures satisfying these requirements. High performance is achieved via pipelining, whereas desired fault tolerance is realized using hardware redundancy. The framework has been used to synthesize high-performance and fault-tolerant microarchitectures for a variety of signal-processing algorithms. Additionally, the framework has been used to explore design trade-offs between high-performance and high-reliability microarchitectures, subject to a maximum area constraint.
AB - Microarchitectural implementations of real-time signal-processing algorithms for mission-critical applications (such as sensors on spacecrafts) are characterized by two unique requirements: real-time processing of a steady stream of input signals, which requires a high-performance implementation such as pipelining, and reliable operation over the mission lifetime, which mandates support for fault tolerance. We relate high-performance and fault-tolerance constraints to chip area and present a methodology for synthesizing area-efficient microarchitectures satisfying these requirements. High performance is achieved via pipelining, whereas desired fault tolerance is realized using hardware redundancy. The framework has been used to synthesize high-performance and fault-tolerant microarchitectures for a variety of signal-processing algorithms. Additionally, the framework has been used to explore design trade-offs between high-performance and high-reliability microarchitectures, subject to a maximum area constraint.
UR - http://www.scopus.com/inward/record.url?scp=0028417203&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0028417203&partnerID=8YFLogxK
U2 - 10.1016/0164-1212(94)90058-2
DO - 10.1016/0164-1212(94)90058-2
M3 - Article
AN - SCOPUS:0028417203
SN - 0164-1212
VL - 25
SP - 73
EP - 84
JO - The Journal of Systems and Software
JF - The Journal of Systems and Software
IS - 1
ER -