HIGH PERFORMANCE COMPUTING
High Performance Computing integrates a set of tools and techniques for distribution and execution of jobs on a supercomputer, which in our case is a cluster of powerful servers linked together through a dedicated network, and which access to a file system using NFS.
Particular libraries such as the Message Pass Interface (MPI) solve execution of jobs in a parallel and distributed way, which define how the processes communicate with each other by sending and receiving messages.
HPC available applications.
Bioinformatics and Life Sciences applications are the most used. The platform allows us to satisfy the processing requirements associated with Next Generation Sequencing (NGS), involving the treatment of large volumes of data that traditional computer systems cannot handle or, if they do, it is at a high cost of processing time.
However, system is able to implement any general-purpose applications, so it is capable of incorporating all type of scientific tool or application.
Taking into account the main research projects developed in BioCubaFarma centers, a set of scientific applications were installed. Among them, we can mention:
- GROMACS (molecular dynamics)
- Rosseta Commons
- FreeSurfer (image processing)
- Anaconda (PYTHON library and packages)
- Weather research and Forecasting (WRF)
Faced with the need of researchers to use a new application, our specialists will install them with the best architecture optimization options and will provide a parallelism profile for evaluation purpose, all in order to achieve more efficient processing in the shortest possible time.
The platform was developed using open and free software, based on most recent CentOS 8 operating system. A resource management system (Sun Grid Engine) was implemented, for planning and orderly dispatch of tasks. A monitoring web page is available (http://ganglia.biocubafarma.cu), in which users can see in real time job status and resources consumption.