An optimization approach for agent-based computational models of biological development. (July 2018)

• Record Type: Journal Article
• Authors: Gonzalez-de-Aledo, Pablo; Vladimirov, Andrey; Manca, Marco; Baugh, Jerry; Asai, Ryo; Kaiser, Marcus; Bauer, Roman
• Is Part Of: Advances in engineering software. Volume 121(2018)
• Page Start: 262
• Page End: 275
• Publication Date: 2018-07
• Subjects: Optimizations -- Parallel computing -- Vectorization -- Coprocessor -- Performance -- Agent-based models -- Biological -- Simulation; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2018.03.010
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: We present a highly parallel implementation of a computer simulation that involves millions of agents interacting in a 3D environment. We study the simulation of biological development using various heterogeneous and multicore platforms. We explain a general approach to transform sequential code of biological dynamics to run on modern, highly parallel architectures such as the Intel Knights Landing, Broadwell, Sand y-Bridge and AMD Opteron. We present the techniques that enabled to obtain more than 500 × speed-up over the mention eded platforms and simulations. The manuscript exemplifies the innovative use of computational strategies and numerical algorithms for large-scale biological problems. Abstract: Current research in the field of computational biology often involves simulations on high-performance computer clusters. It is crucial that the code of such simulations is efficient and correctly reflects the model specifications. In this paper, we present an optimization strategy for agent-based simulations of biological dynamics using Intel Xeon Phi coprocessors, demonstrated by a prize-winning entry of the "Intel Modern Code Developer Challenge" competition. These optimizations allow simulating various biological mechanisms, in particular the simulation of millions of cells, their proliferation, movements and interactions in 3D space. Overall, our results demonstrate a powerful approach to implement and conduct very detailed and large-scale computational simulations for biological research. We also highlight the main difficulties faced when developing such optimizations, in particular the assessment of the simulation accuracy, the dependencies between different optimization techniques and counter-intuitive effects in the speed of the optimized solution. The overall speedup of 595 × shows a good parallel scalability.