NMP ST8 dependable multiprocessor: TRL6 validation — preliminary results

Space exploration, science, and autonomy missions are requiring ever-increasing bandwidth and processing capacity to the extent that the ability to apply high-performance COTS processors for onboard computing in space is becoming a critical need. To date, dependable multiprocessor (DM) technology has been developed as part of NASA's New Millennium Program (NMP) ST8 (space technology 8) project. DM was one of four technologies selected for the ST8 flight experiment. The objective of the NMP ST8 effort is to combine high-performance, SEU-tolerant, COTS-based cluster processing and SEU-tolerant middleware in an architecture and software framework capable of supporting a wide variety of mission applications. The goal of the dependable multiprocessor project is to provide spacecraft/payload processing capability at speeds 10x - 100x of what is available today, enabling heretofore unrealizable levels of science and autonomy. While targeted primarily to allow "soft," state-of-the-art, COTS components to be flown in space, the software-enhanced SEU tolerance techniques described in this paper are applicable to systems implemented with any level of radiation-hardened or radiation-tolerant components. The effectiveness of software-enhanced SEU tolerance techniques is dependent upon the environment, the susceptibility of the components to radiation-induced SEU, SEL, and SEFI effects, the mission, and the application. One condition that obviously negates the potential benefits of software-based SEU enhancement techniques is the use of a component that exhibits catastrophic latch-up. As long as latch-ups are not catastrophic and can be mitigated by cycling power to the component with no damage or degradation of the component, software-based SEU enhancement techniques can be effective. The effectiveness of software-based SEU tolerance techniques may also be limited by the SEE rates of the components. If the SEE rates are so high that the system spends most of its time in recovery, the availability of the system to support mission processing may be low. The highest performance COTS component combined with low system availability may yield a system whose net delivered throughput and throughput density may be low. The DM project held a successful Preliminary TRL6 Technology Validation Review in September 2008. The Final TRL6 Review is currently scheduled for March 2009. This paper provides a brief high-level summary of DM technology and focuses on the TRL6 technology validation criteria and how those criteria have been met to date based on the results presented at the September Review.