Institute of Applied Mathematics and Mechanics NAS of Ukraine, Donetsk PARALLEL FAULT SIMULATION ON MULTI-CORE PROCESSORS

In this paper we propose a fault simulation algorithm that utilizes all cores in multi-core processors. We adapt for multi-core workstation our early proposed distributed fault simulation algorithm. Proposed algorithm uses multi thread execution. The algorithm is based on the well-known «master-slave» approach in which one thread is nominated as a master and controls the calculation on all the other cores of processor. To maximize utilization of the cores a scheme with static fault list partitioning is used. The speed-up coefficient of the simulation time obtained during machine experiments is up to 3.44 times on the quad core system.

[1]  Tassos Markas,et al.  On distributed fault simulation , 1990, Computer.

[2]  Jacob A. Abraham,et al.  Fault simulation in a distributed environment , 1988, 25th ACM/IEEE, Design Automation Conference.Proceedings 1988..

[3]  Chen-Shang Lin,et al.  Hyhope : A Fast Fault Simulator With Efficient Simulation Of Hypertrophic Faults , 1994, IEEE/ACM International Conference on Computer-Aided Design.

[4]  S. Ghosh,et al.  NODIFS: a novel, distributed circuit-partitioning based algorithm for fault simulation of combinational and sequential digital designs on loosely-coupled parallel processors , 1991, Conference Proceedings 1991 IEEE International Conference on Systems, Man, and Cybernetics.

[5]  David Bryan,et al.  Combinational profiles of sequential benchmark circuits , 1989, IEEE International Symposium on Circuits and Systems,.