Abstract
With silicon technology aggressively scaling, the aging-induced long-term reliability becomes the prominent problem of integrated circuits. The reliability of homogeneous multi-core system is improved and the aging is relieved through balancing the workload between cores. While for heterogeneous systems, this will increase the gap of original MTTF differences of each core. By considering the original diversity of MTTF, a MILP-based task scheduling method is proposed to balance the MTTF of whole system. Experimental results show that, compared to the workload balancing algorithms, our approach improve MTTF of system over 30 % on average.
Author supplied keywords
Cite
CITATION STYLE
Liang, H., Dai, Y., Yi, M., Xu, D., & Huang, Z. (2015). MTTF-Aware reliability task scheduling for heterogeneous multicore system. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9529, pp. 716–727). Springer Verlag. https://doi.org/10.1007/978-3-319-27122-4_50
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
