PipeMEM: A Framework to Speed Up BWA-MEM in Spark with Low Overhead

(1) Background: DNA sequence alignment process is an essential step in genome analysis. BWA-MEM has been a prevalent single-node tool in genome alignment because of its high speed and accuracy. The exponentially generated genome data requiring a multi-node solution to handle large volumes of data currently remains a challenge. Spark is a ubiquitous big data platform that has been exploited to assist genome alignment in handling this challenge. Nonetheless, existing works that utilize Spark to optimize BWA-MEM suffer from higher overhead. (2) Methods: In this paper, we presented PipeMEM, a framework to accelerate BWA-MEM with lower overhead with the help of the pipe operation in Spark. We additionally proposed to use a pipeline structure and in-memory-computation to accelerate PipeMEM. (3) Results: Our experiments showed that, on paired-end alignment tasks, our framework had low overhead. In a multi-node environment, our framework, on average, was 2.27× faster compared with BWASpark (an alignment tool in Genome Analysis Toolkit (GATK)), and 2.33× faster compared with SparkBWA. (4) Conclusions: PipeMEM could accelerate BWA-MEM in the Spark environment with high performance and low overhead.

Implementasi Algoritma Negascout Untuk Permainan Checkers

Checker is a zero sum game which means if one player is declared win then the other player is declared lose, if one player gets 1 point then the other player gets -1 point. In this type of game every player have full knowledge of other player condition, like every move the other player has, what kind of pawn and the position of other player's pawn. This game will be implemented with Negascout and compared to Alpha-Beta to see wether it offer better or worse performance than Alpha-Beta. Both algorithm will be given 5 identical board condition to solve and the search depth will be limited to 4, 6 and 8 level. The result showed Negascout outperformed Alpha-Beta on 86% of the test performed. It searched less node than Alpha-Beta especially with depth 8. The result also showed Negascout found one difference node solution compared to Alpha-Beta with identical heuristic score. Negascout returned identical heuristic score with Alpha-Beta on all test. Kata Kunci : Negascout, Permainan, Checkers, Kecerdasan Buatan

A Parallelized, Dynamic Solution Adaptive, Multi-Grid, Overset Chimera Unstructured Grid Solver

This paper presents a parallelized, dynamic solution adaptive, multi-grid, overset Chimera unstructured finite volume solver for all-speed flows. The methods and schemes applicable for all speeds were implemented in the general-purpose CFD computer code. The hanging node solution adaptation, parallel computing algorithm using MPI libraries, overset chimera grid to allow multiple moving bodies calculation and algebraic multi-grid speeding-up convergence method were all facilitated in the solver. With the enhancement of higher order scheme data reconstruction, higher order upwind-biased differencing, multi-pressure-correction, and Bi-CGSTAB and GMRES matrix solvers, the solver can efficiently and accurately handle flow problems in several cases including benchmark and practical.

Maximum Temperatures in Diamond Heat Spreaders Using the Surface Element Method

Diamond heat spreaders are very attractive for lowering the temperature of laser diodes and computer chip components. The use of diamonds as heat spreaders is very advantageous because its thermal conductivity is so large, about four times that of copper. The diamond heat spreader is mounted on a semi-infinite heat sink. The purposes of this paper are to present (a) a surface element method for the analysis of such composite systems, (b) a set of convenient algebraic equations for the maximum temperatures, (c) optimal geometry conditions and (d) some accurate numerical results. The analysis method is an adaptation of the unsteady surface element method but is different because the present problem is a steady-state one. The surface element method with one node gives a relatively simple algebraic solution, which contains all the important dimensionless groups. The one-node solution is simple, accurate, and has a form that can give deep insight into the effects of various parameters; for example, it premits derivation of the optimal geometry corresponding to the minimal temperature at the hot spot. The study shows that the optimal geometry of the thickness to radius aspect ratio of the diamond heat spreaders is about 0.4. In addition to the single-node analysis, a multinode analysis is developed and very accurate results are presented. These results show that the single-node analysis is generally within 2 percent.