In this paper, we propose a scalable parallel algorithm for simulating the cardiac fluid-structure interactions (FSI) of a patient-specific human left ventricle. It provides an efficient forward solver to deal with the induced sub-problems in solving an inverse problem that can be used to quantify the interested parameters. The FSI between the blood flow and the myocardium is described in an arbitrary Lagrangian-Eulerian (ALU) framework, in which the velocity and stress are assumed being continuous across the fluid-structure interface. The governing equations are discretized by using a finite element method and a fully implicit backward Eulerian formula, and the resulting algebraic system is solved by using a parallel Newton-Krylov-Schwarz algorithm. We numerically show that the algorithm is robust with respect to multiple model parameters and scales well up to 2300 processor cores. The ability of the proposed method to produce qualitatively true prediction is also demonstrated via comparing the simulation results with the clinic data.

In this paper, we present a discrete unified gas kinetic scheme (DUGKS) on unstructured grids for high-speed viscid compressible flows on the basis of double distribution function (the density and the total energy distribution functions) Boltzmann-BGK equations. In the DUGKS, the discrete equilibrium distribution functions are constructed based on a D2Q17 circular function. In order to accelerate the simulation, we also illustrate a corresponding parallel algorithm. The DUGKS is validated by two benchmark problems, i.e., flows around the NACA0012 airfoil and flows past a circular cylinder with the Mach numbers range from 0.5 to 2.5. Good agreements with the referenced results are observed from the numerical results. The results of parallel test indicate that the DUGKS is highly parallel scalable, in which the parallel efficiency achieves $93.88\mathrm{\%}$ on a supercomputer using up to $4800$ processors. The proposed method can be utilized for high-resolution numerical simulation of complex and high Mach number flows.

With its wider acceptability, cloud can host a diverse set of data and applications ranging from entertainment to personal to industry. The foundation of cloud computing is based on virtual machines where boundaries among the application data are very thin, and the potential of data leakage exists all the time. For instance, a virtual machine covert timing channel is an aggressive mechanism to leak confidential information through shared components or networks by violating isolation and security policies in practice. The performance of a covert timing channel (covert channel) is crucial to adversaries and attempts have been made to improve the performance of covert timing channels by advancing the encoding mechanism and covert information carriers. Though promising, the redundancy of the covert message is mainly overlooked. This paper applies three encoding schemes namely run-length, Huffman, and arithmetic encoding schemes for data compression of a virtual machine covert timing channel by exploiting redundancy. Accordingly, the paper studies the performance of such channels according to their capacity. Unfortunately, we show that these encoding schemes still contain redundancy in a covert channel scenario, and thereby a new encoding scheme namely optimized Run-length encoding (OptRLE) is presented that greatly enhances the performance of a covert timing channel. Several optimizations schemes adopted by OptRLE are also discussed, and a mathematical model of the behavior of an OptRLE-based covert timing channel is proposed. The theoretical capacity of a channel can be obtained using the proposed model. Our analysis reveals that OptRLE further improves the performance of a covert timing channel, in addition to the effects of the optimizations. Experimental result shows how OptRLE affects the size of covert data and the capacity of covert timing channels, and why the performance of the covert timing channel is improved.