Scientists are enabling every computer to contribute to the research on curing diseases

Thread Content

BOINC, or Berkeley Open Infrastructure for Network Computing, is one of the mainstream distributed computing platforms today; it is a distributed computing system developed by the Department of Computer Science at the University of California, Berkeley. It was originally designed for the SETIhome project, but it has gradually been applied in other fields including mathematics, medicine, astronomy, meteorology, and more. BOINC is currently aimed at providing researchers with the powerful computing power derived from a large number of personal computers around the world. As of March 25, 2012, BOINC had 290,743 volunteers and 472,834 computers active. 24-hour average: 6.300 PetaFLOPS. The operation principle is that personal computers running the BOINC software use their CPUs to perform calculations when they are not in use. Even if a personal computer is in use, BOINC will still be used for calculations if there are available CPU cycles. If the user’s computer is equipped with a graphics processor (GPU) that supports NVIDIA CUDA or ATI Stream technology, the computing speed of certain BOINC projects can be increased by 2 to 10 times compared to versions that use only a CPU (2). When participants use personal computers to work on BOINC projects, BOINC communicates with the project’s servers, which then provide workunits (abbreviated as WUs) to the personal computers, after which the personal computers carry out calculations on those workunits. Once completed, BOINC will upload the results to the project’s server. Each project has its own server, which is used to coordinate the work of the computers involved, including sending work packages, receiving the processed results, and verifying large amounts of data before further processing them into the data required by researchers. Since individual personal computers may make errors during computation, servers generally send the same work package to multiple participants and compare the results. BOINC has a points system that indirectly reflects the contributions of participants. Because the projects that can run on BOINC vary greatly; for example, the workload of Task A may take 3 hours to complete on a particular computer, while the workload of Task B requires 30 hours to be completed on the same computer. Clearly, it is not feasible to measure the workload based on the number of tasks completed ； Similarly, computer performance also varies, and using CPU time to measure workload is even less appropriate. The scoring system can only determine the actual amount of computation completed by participants through certain algorithms. Starting from version 6.4.5, BOINC began to support GPU computing; currently, there are various GPU projects such as GPUgrid.net, SETIhome, Milkywayhome, A.UAhome, Collatz Conjecture, and DNETCHome. In simple terms, it involves downloading a packet-processing program, running it, and then sending the packets back to the main station. For detailed instructions and downloads, please visit the China Distributed Computing Portal. As BOINC has become more widely used, an increasing number of research projects rely on it. Below is a list of projects categorized by function: (**Bold** indicates **important projects**). Biology, Medicine: DockingHome – This project focuses on studying the atomic-level structure and details of protein binding and reactions, and utilizes these findings to develop drugs for treating human diseases. DrugDiscoveryHome — Developing drugs to treat human diseases. Malaria Control — Simulates the effects and control of malaria. Predictorhome — predicts protein structures; it is no longer in operation. Proteinshome — inferred the sequence of DNA; it is no longer in operation. GPUGRID.net — Conducts research related to molecular biomechanics, primarily using NVidia GPUs that support CUDA. Rosettahome — studying protein folding. RALPHHome — a test project for RosettaHome. SIMAP (Similarity Matrix of Proteins) – a database for calculating protein homology sequences, offering professional search tools for these sequence data. TANPAKU—a method for calculating protein structures using Brownian dynamics—has been discontinued. POEMHOME — Using energy methods to study protein structure. MindModelingHome — the cognitive science of the human brain. SuperlinkTechnion — helps scientists study human genes and the diseases caused by their abnormalities. The Lattice Project — a distributed computing platform for research in bioinformatics at the University of Maryland in the United States. Virtual Prairie CelsHome RNA world — Research and analysis of the molecular structure of RNA in organisms. DNAHome—studies the gene regulatory role of DNA in organisms. Currently, many projects around the world that could benefit humanity are stalled due to insufficient computing power, or fail to be completed promptly! Therefore, I urge everyone to make use of the idle computing power of CPUs to carry out calculations, as a way to contribute to humanity! By incorporating distributed computing, you can contribute to these projects. Everyone can visit the ****** China Distributed Computing Portal****** to download the client, select the projects that interest them, and once they join, tasks will be downloaded automatically, calculations will be carried out automatically, and results will be uploaded automatically. Moreover, the priority for these calculations is very low; in other words, calculations are performed only when the CPU is free. If you have your own tasks, the CPU will prioritize those tasks first, so the computer won’t slow down at all. It’s also possible to manually adjust the CPU usage. The results produced by BOINC and Folding.home are publicly available. Many biological and medical research projects in China make use of the findings from Folding.home. For example: “Design and Implementation of a Drug Discovery Grid” by Zhang Wenju (1), Chen Shudong (1), Liu Liao (3), Ma Fanyuan (1), Shen Jianhua (2); (1) Department of Computer Science, Shanghai Jiao Tong University, Shanghai 200030; (2) Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203; (3) Jiangnan Institute of Computing Technology, Wuxi 214083. “A Distributed Grid Computing Framework and Its Application in Large-Scale Molecular Dynamics Simulations” by Wang Wenrui (1), Chen Guoliang (1), Xing Lirong (2), Chen Huaping (1), Sun Guangzhong (1), Shan Jiulong (1); (1) **High-Performance Computing Center**, University of Science and Technology of China, Hefei 230027; (2) Department of Mathematics and Physics, Shanghai University of Electric Power, Shanghai 201300. “Parallel Optimization Algorithm for GROMOS96 Molecular Dynamics Simulations” by Wang Wenrui, Chen Guoliang, Sun Guangzhong, **High-Performance Computing Center**, Department of Computer Science, University of Science and Technology of China, Hefei 230027