DocumentCode :
2289433
Title :
Radiation fault modeling and fault rate estimation for a COTS based space-borne supercomputer
Author :
Karapetian, Arbi V. ; Some, Raphael R. ; Beahan, John J.
Author_Institution :
Jet Propulsion Lab., California Inst. of Technol., Pasadena, CA, USA
Volume :
5
fYear :
2002
fDate :
2002
Firstpage :
80841
Abstract :
Development of the Remote Exploration and Experimentation (REE) Commercial Off The Shelf (COTS) based space-borne supercomputer requires a detailed model of Single Event Upset (SEU) induced faults and fault-effects. Extensive ground based radiation testing has been performed on several generations of the Power PC processor family and related components. A set of relevant environments for NASA missions have been analyzed and detailed. Combining radiation test data, environmental data and architectural analysis, we have developed a radiation fault model for the REE system. The fault model is hierarchically organized and includes scaling factors and optional parameters for fault prediction in future technologies and alternative architectures. It has been implemented in a generic tool, which allows for ease of input and straight forward porting. The model currently includes the Power PC750 (G3), PCI bridge chips, L2 cache SRAM, main memory DRAM, and the Myrinet packet switched network. In this paper, we present the REE radiation fault model and accompanying tool set. We explain its derivation, its structure and use, and the work being done to validate it.
Keywords :
aerospace computing; fault diagnosis; modelling; parallel machines; performance evaluation; radiation effects; reliability; space vehicle electronics; COTS based space-borne supercomputer; L2 cache SRAM; Myrinet packet switched network; NASA missions; PCI bridge chips; Power PC processor family; Power PC750; Remote Exploration Experimentation supercomputer; SEU induced faults; architectural analysis; environmental data; fault prediction; fault rate estimation; main memory DRAM; radiation fault model; radiation test data; scaling factors; single event upset; Data analysis; NASA; Performance evaluation; Power generation; Power system modeling; Predictive models; Single event upset; Space technology; Supercomputers; System testing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Aerospace Conference Proceedings, 2002. IEEE
Print_ISBN :
0-7803-7231-X
Type :
conf
DOI :
10.1109/AERO.2002.1035378
Filename :
1035378
Link To Document :
بازگشت