Autotuning Parallel Programs by Model Checking

The paper presents a new approach to autotuning data-parallel programs. Autotuning is a search for optimal program settings which maximize its performance. The novelty of the approach lies in the use of the model checking method to find the optimal tuning parameters by the method of counterexamples. In our work, we abstract from specific programs and specific processors by defining their representative abstract patterns. Our method of counterexamples implements the following four steps. At the first step, an execution model of an abstract program on an abstract processor is described in the language of a model checking tool. At the second step, in the language of the model checking tool, we formulate the optimality property that depends on the constructed model. At the third step, we find the optimal values of the tuning parameters by using a counterexample constructed during the verification of the optimality property. In the fourth step, we extract the information about the tuning parameters from the counter-example for the optimal parameters. We apply this approach to autotuning parallel programs written in OpenCL, a popular modern language that extends the C language for programming both standard multi-core processors (CPUs) and massively parallel graphics processing units (GPUs). As a verification tool, we use the SPIN verifier and its model representation language Promela, whose formal semantics is good for modelling the execution of parallel programs on processors with different architectures.

On the L∞-maximization of the solution of Poisson's equation: Brezis–Gallouet–Wainger type inequalities and applications

For the solution of the Poisson problem with an L∞ right hand side \begin{cases} -\Delta u(x) = f (x) & {\rm in}\ D, \\ u=0 & {\rm on}\ \partial D \end{cases} we derive an optimal estimate of the form \|u\|_\infty\leq \|f\|_\infty \sigma_D(\|f\|_1/\|f\|_\infty), where σ D is a modulus of continuity defined in the interval [0, |D|] and depends only on the domain D. The inequality is optimal for any domain D and for any values of $\|f\|_1$ and $\|f\|_\infty .$ We also show that \sigma_D(t)\leq\sigma_B(t),\text{ for }t\in[0,|D|], where B is a ball and |B| = |D|. Using this optimality property of σ D , we derive Brezis–Galloute–Wainger type inequalities on the L∞ norm of u in terms of the L1 and L∞ norms of f. As an application we derive L∞ − L1 estimates on the k-th Laplace eigenfunction of the domain D.

Secure Communication for Two-way Relay Networks with Imperfect CSI

This paper considers a two-way relay network, where two source nodes exchange messages through several relays in the presence of an eavesdropper, and the channel state information (CSI) of the eavesdropper is imperfectly known. The amplify-and-forward relay protocol is used and the relay beamforming weights are designed. The model is built up to minimize the total relay transmit power while guaranteeing the quality of service at users and preventing the eavesdropper from decoding the signals. Due to the imperfect CSI, a semi-infinite programming problem is obtained. An algorithm is proposed to solve the problem, and the iterative points are updated through the linesearch technique, where the feasibility are preserved during iterations. The optimality property is analyzed. The obtained subproblems are quadratic constrained quadratic programming problems, either with less than $4$ constraints or with only one variable, which are solved optimally. Simulation results demonstrate the importance of the proposed model, and imply that the proposed algorithm is efficient and converges very fast, where more than 85% of the problems are solved optimally.

From snapshots to modal expansions – bridging low residuals and pure frequencies

Data-driven low-order modelling has been enjoying rapid advances in fluid mechanics. Arguably, Sirovich (Q. Appl. Maths, vol. XLV, 1987, pp. 561–571) started these developments with snapshot proper orthogonal decomposition, a particularly simple method. The resulting reduced-order models provide valuable insights into flow physics, allow inexpensive explorations of dynamics and operating conditions, and enable model-based control design. A winning argument for proper orthogonal decomposition (POD) is the optimality property, i.e. the guarantee of the least residual for a given number of modes. The price is unpleasant frequency mixing in the modes which complicates their physical interpretation. In contrast, temporal Fourier modes and dynamic mode decomposition (DMD) provide pure frequency dynamics but lose the orthonormality and optimality property of POD. Sieber et al. (J. Fluid Mech., vol. 792, 2016, pp. 798–828) bridge the least residual and pure frequency behaviour with an ingenious interpolation, called spectral proper orthogonal decomposition (SPOD). This article puts the achievement of the TU Berlin authors in perspective, illustrating the potential of SPOD and the challenges ahead.