Abstract. Science and engineering problems frequently require solving a sequence of dual linear systems. Besides having to store only few Lanczos vectors, using the BiConjugate Gradient method (BiCG) to solve dual linear systems has advantages for specific applications. For example, using BiCG to solve the dual linear systems arising in interpolatory model reduction provides a backward error formulation in the model reduction framework. Using BiCG to evaluate bilinear forms -for example, in quantum Monte Carlo (QMC) methods for electronic structure calculations -leads to a quadratic error bound. Since our focus is on sequences of dual linear systems, we introduce recycling BiCG, a BiCG method that recycles two Krylov subspaces from one pair of dual linear systems to the next pair. The derivation of recycling BiCG also builds the foundation for developing recycling variants of other bi-Lanczos based methods, such as CGS, BiCGSTAB, QMR, and TFQMR.We develop an augmented bi-Lanczos algorithm and a modified two-term recurrence to include recycling in the iteration. The recycle spaces are approximate left and right invariant subspaces corresponding to the eigenvalues closest to the origin. These recycle spaces are found by solving a small generalized eigenvalue problem alongside the dual linear systems being solved in the sequence.We test our algorithm in two application areas. First, we solve a discretized partial differential equation (PDE)

Abstract. Krylov subspace recycling is a process for accelerating the convergence of sequences of linear systems. Based on this technique, the recycling BiCG algorithm has been developed recently. Here, we now generalize and extend this recycling theory to BiCGSTAB. Recycling BiCG focuses on efficiently solving sequences of dual linear systems, while the focus here is on efficiently solving sequences of single linear systems (assuming non-symmetric matrices for both recycling BiCG and recycling BiCGSTAB).As compared with other methods for solving sequences of single linear systems with non-symmetric matrices (e.g., recycling variants of GMRES), BiCG based recycling algorithms, like recycling Bi-CGSTAB, have the advantage that they involve a short-term recurrence, and hence, do not suffer from storage issues and are also cheaper with respect to the orthogonalizations.We modify the BiCGSTAB algorithm to use a recycle space, which is built from left and right approximate invariant subspaces. Using our algorithm for a parametric model order reduction example gives good results. We show about 40% savings in the number of matrix-vector products and about 35% savings in runtime.

In the field programmable gate array (FPGA) design flow, one of the most time-consuming steps is the routing of nets. Therefore, there is a need to accelerate it. In a recent work by Hoo et al., the authors have developed a linear programming (LP)-based framework that parallelizes this routing process to achieve significant speed-ups (the resulting algorithm is termed as ParaLaR). However, this approach has certain weaknesses. Namely, the constraints violation by the solution and a standard routing metric could be improved. We address these two issues here. In this paper, we use the LP framework of ParaLaR and solve it using the primal-dual sub-gradient method that better exploits the problem properties. We also propose a better way to update the size of the step taken by this iterative algorithm. We call our algorithm as ParaLarPD. We perform experiments on a set of standard benchmarks, where we show that our algorithm outperforms not just ParaLaR but the standard existing algorithm VPR as well. We perform experiments with two different configurations. We achieve 20% average improvement in the constraints violation and the standard metric of the minimum channel width (both of which are related) when compared with ParaLaR. When compared to VPR, we get average improvements of 28% in the minimum channel width (there is no constraints violation in VPR). We obtain the same value for the total wire length as by ParaLaR, which is 49% better on an average than that obtained by VPR. This is the original metric to be minimized, for which ParaLaR was proposed. Next, we look at the third and easily measurable metric of critical path delay. On an average, ParaLarPD gives 2% larger critical path delay than ParaLaR and 3% better than VPR. We achieve maximum relative speed-ups of up to seven times when running a parallel version of our algorithm using eight threads as compared to the sequential implementation. These speed-ups are similar to those as obtained by ParaLaR.CAD (computer-aided design) tools. This can be achieved in two ways. First, by parallelizing the routing algorithms for hardware having multiple cores. However, the pathfinder algorithm [3], which is one of the most commonly used FPGA routing algorithm is intrinsically sequential. Hence, this approach seems inappropriate for parallelizing all types of FPGA routing algorithms.Second, instead of compiling the entire design together, the users can partition their design, compile partitions progressively, and then assemble all the partitions to form the entire design. Some existing works have proposed this approach [4,5]. However, the routing resources required by one partition may be held by another partition, i.e., there is no guarantee to have balanced partitions. In other words, in this approach, there is a need to tackle the difficulties arising in sharing of routing resources.The authors in ParaLaR [6] overcome the limitations of existing approaches by formulating the FPGA routing problem as an optimization problem [7]. Here, the objective function is linear an...

Please cite this article in press as: A. Amritkar et al., Recycling Krylov subspaces for CFD applications and a new hybrid recycling solver, J. Comput. Phys. (2015), http://dx. AbstractWe focus on robust and efficient iterative solvers for the pressure Poisson equation in incompressible Navier-Stokes problems. Preconditioned Krylov subspace methods are popular for these problems, with BiCGStab and GMRES(m) most frequently used for nonsymmetric systems. BiCGStab is popular because it has cheap iterations, but it may fail for stiff problems, especially early on as the initial guess is far from the solution. Restarted GMRES is better, more robust, in this phase, but restarting may lead to very slow convergence. Therefore, we evaluate the rGCROT method for these systems. This method recycles a selected subspace of the search space (called recycle space) after a restart. This generally improves the convergence drastically compared with GMRES(m). Recycling subspaces is also advantageous for subsequent linear systems, if the matrix changes slowly or is constant. However, rGCROT iterations are still expensive in memory and computation time compared with those of BiCGStab. Hence, we propose a new, hybrid approach that combines the cheap iterations of BiCGStab with the robustness of rGCROT. For the first few time steps the algorithm uses rGCROT and builds an effective recycle space, and then it recycles that space in the rBiCGStab solver.We evaluate rGCROT on a turbulent channel flow problem, and we evaluate both rGCROT and the new, hybrid combination of rGCROT and rBiCGStab on a porous medium flow problem. We see substantial performance gains on both problems.

Abstract. Quantum Monte Carlo (QMC) methods are often used to calculate properties of many body quantum systems. The main cost of many QMC methods, for example the variational Monte Carlo (VMC) method, is in constructing a sequence of Slater matrices and computing the ratios of determinants for successive Slater matrices. Recent work has improved the scaling of constructing Slater matrices for insulators so that the cost of constructing Slater matrices in these systems is now linear in the number of particles, whereas computing determinant ratios remains cubic in the number of particles. With the long term aim of simulating much larger systems, we improve the scaling of computing the determinant ratios in the VMC method for simulating insulators by using preconditioned iterative solvers.The main contribution of this paper is the development of a method to efficiently compute for the Slater matrices a sequence of preconditioners that make the iterative solver converge rapidly. This involves cheap preconditioner updates, an effective reordering strategy, and a cheap method to monitor instability of ILUTP preconditioners. Using the resulting preconditioned iterative solvers to compute determinant ratios of consecutive Slater matrices reduces the scaling of QMC algorithms from O(n 3 ) per sweep to roughly O(n 2 ), where n is the number of particles, and a sweep is a sequence of n steps, each attempting to move a distinct particle. We demonstrate experimentally that we can achieve the improved scaling without increasing statistical errors. Our results show that preconditioned iterative solvers can dramatically reduce the cost of VMC for large(r) systems.

Multi-kernel learning has been well explored in the recent past and has exhibited promising outcomes for multi-class classification and regression tasks. In this paper, we present a multiple kernel learning approach for the One-class Classification (OCC) task and employ it for anomaly detection. Recently, the basic multi-kernel approach has been proposed to solve the OCC problem, which is simply a convex combination of different kernels with equal weights. This paper proposes a Localized Multiple Kernel learning approach for Anomaly Detection (LMKAD) using OCC, where the weight for each kernel is assigned locally. Proposed LMKAD approach adapts the weight for each kernel using a gating function. The parameters of the gating function and one-class classifier are optimized simultaneously through a two-step optimization process. We present the empirical results of the performance of LMKAD on 25 benchmark datasets from various disciplines. This performance is evaluated against existing Multi Kernel Anomaly Detection (MKAD) algorithm, and four other existing kernel-based one-class classifiers to showcase the credibility of our approach. LMKAD achieves significantly better Gmean scores while using a lesser number of support vectors compared to MKAD. Friedman test is also performed to verify the statistical significance of the results claimed in this paper.

Dynamical systems are pervasive in almost all engineering and scientific applications. Simulating such systems is computationally very intensive. Hence, Model Order Reduction (MOR) is used to reduce them to a lower dimension. Most of the MOR algorithms require solving large sparse sequences of linear systems. Since using direct methods for solving such systems does not scale well in time with respect to the increase in the input dimension, efficient preconditioned iterative methods are commonly used. In one of our previous works, we have shown substantial improvements by reusing preconditioners for the parametric MOR (Singh et al. 2019). Here, we had proposed techniques for both, the non-parametric and the parametric cases, but had applied them only to the latter. We have three main contributions here. First, we demonstrate that preconditioners can be reused more effectively in the non-parametric case as compared to the parametric one. Second, we show that reusing preconditioners is an art via detailed algorithmic implementations in multiple MOR algorithms. Third and final, we demonstrate that reusing preconditioners for reducing a real-life industrial problem (of size 1.2 million), leads to relative savings of up to 64% in the total computation time (in absolute terms a saving of 5 days).

Models coming from different physical applications are very large in size. Simulation with such systems is expensive so one usually obtains a reduced model (by model reduction) that replicates the input-output behaviour of the original full model. A recently proposed algorithm for model reduction of bilinear dynamical systems, Bilinear Iterative Rational Krylov Algorithm (BIRKA), does so in a locally optimal way. This algorithm requires solving very large linear systems of equations. Usually these systems are solved by direct methods (e.g., LU), which are very expensive. A better choice is iterative methods (e.g., Krylov). However, iterative methods introduce errors in linear solves because they are not exact. They solve the given linear system up to a certain tolerance. We prove that under some mild assumptions BIRKA is stable with respect to the error introduced by the inexact linear solves. We also analyze the accuracy of the reduced system obtained from using these inexact solves and support all our results by numerical experiments.