Parallel Computing and Optimization Techniques

Sparse Linear Algebra Accelerates Genotype Representation Graph Computation at Biobank Scale

This paper introduces Mikado, a method using sparse linear algebra to accelerate genotype representation graph computations, significantly improving speed over existing tools on large biobank datasets.

Revisiting CPUs for Protein Folding: Xeon-Based Acceleration of AlphaFold2

This paper presents an optimized AlphaFold2 implementation for CPUs, achieving significant speedups and higher throughput compared to GPU-based solutions.

DistPCA: Tera-Scale Genomic PCA via Out-of-Core Distributed Parallelism

DistPCA improves genomic PCA for large datasets by enabling distributed, out-of-core processing with high scalability and efficiency.

GPU acceleration of cell-based simulations in Chaste using FLAME GPU 2

This paper integrates FLAME GPU 2 with Chaste to accelerate cell-based simulations, achieving significant speedups and improved mechanical modeling accuracy.

GROMODEX: Optimisation of GROMACS Performance through a Design of Experiment Approach

GROMODEX optimizes GROMACS MD simulations using a Design of Experiments approach, improving performance across different hardware and molecular systems.

GAPiM: Discovering Genetic Variations on a Real Processing-in-Memory System

The paper implements and evaluates a modified Pair-HMM algorithm on a PiM platform for faster variant calling in genome analysis.

Clustering the Planet: An Exascale Approach to Determining Global Climatype Zones

The paper uses exascale computing to generate high-resolution global geoclimate zones by clustering environmental similarity data from 71 variables over 50 years.

A Multithreaded Model for Cancer Tissue Heterogeneity: An Application

This paper presents a multithreaded CPU method to run GPU-compatible models for cancer tissue heterogeneity, showing it is faster than GPU implementations on real experimental data.

Accelerating genomic workflows using NVIDIA Parabricks

This study benchmarks NVIDIA Parabricks, a GPU-accelerated software suite, showing significant speedups for germline variant calling and cost savings on cloud platforms.

Cross-comparison of state of the art neuromorphological simulators on modern CPUs and GPUs using the Brain Scaffold Builder

The paper benchmarks two neuromorphological simulators, NEURON and Arbor, on CPUs and GPUs, showing Arbor can be 60-400x faster and 100x more energy efficient.

Variant Calling Parallelization on Processor-in-Memory Architecture

The paper introduces a PIM architecture to accelerate variant calling by reducing data movement and energy consumption, achieving significant speed-up and cost savings compared to software, FPGA, and GPU solutions.

BigMPI4py: Python module for parallelization of Big Data objects

BigMPI4py enables efficient parallelization of large data objects in Python, validated on DNA methylation data to identify differentially methylated CpGs across germ layers.

Parallel and Scalable Precise Clustering for Homologous Protein Discovery

This paper introduces ClusterMerge, a precise clustering algorithm that efficiently identifies homologous proteins by reducing computational complexity and enabling parallel processing.

Spike: A GPU Optimised Spiking Neural Network Simulator

The paper introduces Spike, a GPU-optimized spiking neural network simulator with optimizations that significantly improve performance over CPU-based simulators.

Brian2GeNN: a system for accelerating alarge variety of spiking neural networkswith graphics hardware

Brian2GeNN accelerates spiking neural network simulations by translating Brian scripts into GPU-accelerated C++ code via GeNN, improving performance by tens to hundreds of times.

Sequence Alignment Through the Looking Glass

This study analyzes how sequence aligners use modern processors and finds they underutilize hardware, identifying bottlenecks for future aligner design.

Benchmarking BarraCUDA on Epigenetic DNA and nVidia Pascal GPUs

The paper benchmarks BarraCUDA on epigenetic DNA data using nVidia Pascal GPUs, showing comparable performance to Tesla K40 cards.