The Strong Scaling Advantage of FPGAs in HPC for N-body Simulations

N-body methods are one of the essential algorithmic building blocks of high-performance and parallel computing. Previous research has shown promising performance for implementing n-body simulations with pairwise force calculations on FPGAs. However, to avoid challenges with accumulation and memory access patterns, the presented designs calculate each pair of forces twice, along with both force sums of the involved particles. Also, they require large problem instances with hundreds of thousands of particles to reach their respective peak performance, limiting the applicability for strong scaling scenarios. This work addresses both issues by presenting a novel FPGA design that uses each calculated force twice and overlaps data transfers and computations in a way that allows to reach peak performance even for small problem instances, outperforming previous single precision results even in double precision, and scaling linearly over multiple interconnected FPGAs. For a comparison across architectures, we provide an equally optimized CPU reference, which for large problems actually achieves higher peak performance per device, however, given the strong scaling advantages of the FPGA design, in parallel setups with few thousand particles per device, the FPGA platform achieves highest performance and power efficiency.

Computational complexity continuum within Ising formulation of NP problems

A promising approach to achieve computational supremacy over the classical von Neumann architecture explores classical and quantum hardware as Ising machines. The minimisation of the Ising Hamiltonian is known to be NP-hard problem yet not all problem instances are equivalently hard to optimise. Given that the operational principles of Ising machines are suited to the structure of some problems but not others, we propose to identify computationally simple instances with an ‘optimisation simplicity criterion’. Neuromorphic architectures based on optical, photonic, and electronic systems can naturally operate to optimise instances satisfying this criterion, which are therefore often chosen to illustrate the computational advantages of new Ising machines. As an example, we show that the Ising model on the Möbius ladder graph is ‘easy’ for Ising machines. By rewiring the Möbius ladder graph to random 3-regular graphs, we probe an intermediate computational complexity between P and NP-hard classes with several numerical methods. Significant fractions of polynomially simple instances are further found for a wide range of small size models from spin glasses to maximum cut problems. A compelling approach for distinguishing easy and hard instances within the same NP-hard class of problems can be a starting point in developing a standardised procedure for the performance evaluation of emerging physical simulators and physics-inspired algorithms.

Tackling the rich vehicle routing problem with nature-inspired algorithms

In the last decades, the classical Vehicle Routing Problem (VRP), i.e., assigning a set of orders to vehicles and planning their routes has been intensively researched. As only the assignment of order to vehicles and their routes is already an NP-complete problem, the application of these algorithms in practice often fails to take into account the constraints and restrictions that apply in real-world applications, the so called rich VRP (rVRP) and are limited to single aspects. In this work, we incorporate the main relevant real-world constraints and requirements. We propose a two-stage strategy and a Timeline algorithm for time windows and pause times, and apply a Genetic Algorithm (GA) and Ant Colony Optimization (ACO) individually to the problem to find optimal solutions. Our evaluation of eight different problem instances against four state-of-the-art algorithms shows that our approach handles all given constraints in a reasonable time.

Solving the Cubic Cell Formation Problem Using Simulated Annealing

The Cubic Cell Formation Problem (CCFP) in cellular manufacturing systems consists in decomposing a production system into a set of manufacturing cells, and assigning workers to cells besides parts and machines. The major objective is to obtain manageable cells. Manageable cells mean cells with a minimum value of inter-cell moves of parts and workers and a minimum value of heterogeneity within cells. In this paper, a solution methodology based on a modified simulated annealing heuristic with a proposed neighbourhood search procedure is proposed. The methodology allows building multiple configurations by giving to the decision-maker the ability to control some parameters. Experimental results show that the proposed algorithm gives a promising performance for all problem instances found in the literature.

Applying Metaheuristics to Minimize Work-Related Musculoskeletal Disorders

This article covers the topic of planning and organization of work, which is one of the biggest problems is to establish the most appropriate allocations between human and technical resources, according to the characteristics that define and characterize each individual. These adjustments to decision-making regarding the characteristics of a new larger workforce is a challenge for human resource managers and researchers working to provide well-being and quality of life improvements for employees. The problem of work-related musculoskeletal disorders, coupled with the aging of the active population, may increase the number of citizens with permanent disabilities. Given the complexity and uniqueness of the problems, a decision support system that uses some metaheuristic approaches is presented. The result is a hybrid approach that gives the best solution according to several parameters defined by the decision-maker. Computational results of real problem instances are presented, proving that in most cases, the optimal solution is achieved.

A genetic algorithm for vehicle routing in logistic networks with practical constraints

We optimise a postal delivery problem with time and capacity constraints imposed on vehicles and nodes of the logistic network. Time constraints relate to the duration of routes, whereas capacity constraints concern technical characteristics of vehicles and postal operation outlets. We consider a method which can be applied to a brownfield scenario, in which capacities of outlets can be relaxed and prospective hubs identified. As a solution, we apply a genetic algorithm and test its properties both in small case studies and in a simulated problem instance of a larger (i.e. comparable with real-world instances) size. We show that the genetic operators we employ are capable of switching between solutions based on direct origin-to-destination routes and solutions based on transfer connections, depending on what is more beneficial in a given problem instance. Moreover, the algorithm correctly identifies cases in which volumes should be shipped directly, and those in which it is optimal to use transfer connections within a single problem instance, if an instance in question requires such a selection for optimality. The algorithm is thus suitable for determining hubs and satellite locations. All considerations presented in this paper are motivated by real-life problem instances experienced by the Polish Post, the largest postal service provider in Poland, in its daily plans of delivering postal packages, letters and pallets.

Single Allocation Hub Location with Heterogeneous Economies of Scale

The economies of scale in hub location is usually modeled by a constant parameter, which captures the benefits companies obtain through consolidation. In their article “Single allocation hub location with heterogeneous economies of scale,” Rostami et al. relax this assumption and consider hub-hub connection costs as piecewise linear functions of the flow amounts. This spoils the triangular inequality property of the distance matrix, making the classical flow-based model invalid and further complicates the problem. The authors tackle the challenge by building a mixed-integer quadratically constrained program and by developing a methodology based on constructing Lagrangian function, linear dual functions, and specialized polynomial-time algorithms to generate enhanced cuts. The developed method offers a new strategy in Benders-type decomposition through relaxing a set of complicating constraints in subproblems when such relaxation is tight. The results confirm the efficacy of the solution methods in solving large-scale problem instances.

Transfer Learning for Operator Selection: A Reinforcement Learning Approach

In the past two decades, metaheuristic optimization algorithms (MOAs) have been increasingly popular, particularly in logistic, science, and engineering problems. The fundamental characteristics of such algorithms are that they are dependent on a parameter or a strategy. Some online and offline strategies are employed in order to obtain optimal configurations of the algorithms. Adaptive operator selection is one of them, and it determines whether or not to update a strategy from the strategy pool during the search process. In the filed of machine learning, Reinforcement Learning (RL) refers to goal-oriented algorithms, which learn from the environment how to achieve a goal. On MOAs, reinforcement learning has been utilised to control the operator selection process. Existing research, however, fails to show that learned information may be transferred from one problem-solving procedure to another. The primary goal of the proposed research is to determine the impact of transfer learning on RL and MOAs. As a test problem, a set union knapsack problem with 30 separate benchmark problem instances is used. The results are statistically compared in depth. The learning process, according to the findings, improved the convergence speed while significantly reducing the CPU time.

Minimizing setups and waste when printing labels of consumer goods

A real-world planning problem of a printing company is presented where different sorts of a consumer goods’ label are printed on a roll of paper with sufficient length. The printer utilizes a printing plate to always print several labels of same size and shape (but possibly different imprint) in parallel on adjacent lanes of the paper. It can be decided which sort is printed on which (lane of a) plate and how long the printer runs using a single plate. A sort can be assigned to several lanes of the same plate, but not to several plates. Designing a plate and installing it on the printer incurs fixed setup costs. If more labels are produced than actually needed, each surplus label is assumed to be “scrap”. Since demand for the different sorts may be heterogeneous and since the number of sorts is usually much higher than the number of lanes, the problem is to build “printing blocks”, i.e., to decide how many and which plates to design and how long to run the printer with a certain plate so that customer demand is satisfied with minimum costs for setups and scrap. This industrial application is modeled as an extension of a so-called job splitting problem which is solved exactly and by various decomposition heuristics, partly basing on dynamic programming. Numerical tests compare both approaches with further straightforward heuristics and demonstrate the benefits of decomposition and dynamic programming for large problem instances.

Crew Assignment with Duty Time Limits for Transport Services: Tight Multicommodity Models

Modeling Crew Assignments for Urban Transport Services Using Differentiated Flows Public transit agencies need to judiciously deploy their limited crew members to operate numerous daily scheduled services, while meeting duty and working time regulations for each crew member. Since crew costs account for a large portion of the organizations’ operating expenses, minimizing the total crew and transfer costs is very important. But, with hundreds of daily trips and millions of possible crew itineraries, optimizing trip-to-crew assignment decisions is challenging. In “Crew Assignment with Duty Time Limits for Transport Services: Tight Multicommodity Models,” Balakrishnan, Mirchandani, and Lin propose a novel integer optimization model that represents itineraries as multicommodity flows, differentiated by first trip and depot, to capture the duty time limits and incorporate additional requirements such as selecting equitable schedules. The authors show that this compact model can be tighter than previous formulations, further strengthen the model, and propose a restricted optimization approach combined with an optimality test to generate near-optimal solutions quickly. Extensive computational tests using well-known and real-life problem instances show that the proposed model and solution approach can be very effective in practice.