E-HPC (Elastic High-Performance Computing) is a high-performance computing service with excellent performance, stability, reliability, and elastic expansion provided by Alibaba Cloud. Elastic high-performance computing accumulates computing power and uses parallel computing to solve larger-scale scientific, engineering and commercial problems. It is widely used in scientific research institutions, oil exploration, financial markets, weather forecasting, biopharmaceuticals, gene sequencing, image processing and other industries Wide range of applications.
1. Product advantages
Compared with other HPC clusters, Alibaba Cloud Elastic High Performance Computing E-HPC products feature flexible deployment, elastic resources, data security, high availability, and visibility of results.
(1) Flexible deployment
You can quickly create a high-performance computing cluster on the E-HPC console. One-click deployment of the required high-performance computing environment and application software helps you quickly create a cluster with excellent processing power. After the creation is complete, you can submit a job to start computing.
(2) Elastic resources
According to demand or task queue usage, E-HPC automatically identifies the load of nodes and dynamically adjusts the resources on the cloud. When there are no jobs queued, it will automatically scale down to the minimum number of nodes configured. When a job is submitted to the scheduler queue, a corresponding number of nodes will be automatically created for calculation, making full and efficient use of resources.
(3) Data security
Based on the network access isolation implemented by the VPC, the cluster nodes in the VPC use the security group firewall to perform Layer 3 network access control. The security of the cluster network is fully guaranteed.
Cluster data is stored in Alibaba Cloud File Storage NAS. The NAS’s transmission encryption and storage encryption features are used to ensure that cluster data is not stolen or tampered with. NAS data is stored in multiple copies at the back end, which can effectively reduce data security risks.
(4) High availability
E-HPC cluster nodes are based on cloud server ECS, supercomputing cluster SCC and GPU cloud server, which greatly improves the availability of the cluster.
(5) The results are visible
E-HPC provides visualization service functions, through which you can convert calculation results into readable graphical data. For example, you can directly view the rendered animation effect, or directly view the dynamic graph of weather changes.
2. Application scenarios
(1) Animation rendering
E-HPC supports large-scale multi-machine concurrent services, and is used for graphics rendering in the film and television animation industry.
(2) Weather forecast
E-HPC combines numerical model calculation and analysis of meteorological data and environmental data, and can predict meteorological information such as weather and environment.
(3) Energy exploration
E-HPC can help the exploration industry to analyze exploration data, analyze and simulate the geological structure of the exploration area, so as to accurately find the location of resources.
(4) Life Sciences
Bioinformatics: Use E-HPC to sequence a large number of biological genomes to obtain genomic information and data analysis results to solve difficult problems in the field of biology and medicine.
Dynamics simulation: Use E-HPC to perform large-scale molecular dynamics simulations to predict and analyze the interactions and changes between biological protein molecules and lipid molecules.
New drug development: E-HPC helps developers achieve rapid concurrent processing of a large number of small molecule libraries.
(5) Scientific research education
E-HPC can provide supercomputing services to the government, universities and supercomputing centers for numerical simulation and simulation verification in the research process. Allowing educational experts and scientific researchers to focus on the basic research of the discipline, saving the time of learning and mastering the knowledge of processors and supercomputing.
(6) Simulation manufacturing
The manufacturing industry can use high-performance and elastic computing clusters to assist in the analysis of complex engineering structures and mechanical structures, and optimize product structure and performance through massive data simulations. The simulation manufacturing involved in E-HPC includes smart cars, aerospace, mechanical construction and so on.