Abstract: From more than 500 enterprises across the country to apply for trial, combined with customer experience needs and feedback, continuous improvement and polishing, elastic HIGH performance computing business edition has been greatly optimized, in product performance and experience have a new upgrade.
Click here to view the original text:click.aliyun.com/m/41646/
Elastic high Performance Computing (E-HPC), based on Ali Cloud infrastructure, provides users with one-stop public cloud HPC services. It provides a fast, elastic, secure and interoperable technical computing cloud platform for education, scientific research, enterprises and individuals. is suitable for customers in the high-performance computing industry, including but not limited to industrial design and manufacturing, industrial simulation, weather forecasting, computational chemistry, life sciences, material mechanics, molecular dynamics, astrology, fluid mechanics, nanomaterials, electronic design, data analysis, artificial intelligence, etc. Suitable for scientific research institutes, universities, enterprise researchers, high-performance computing cluster operation and maintenance management engineers, platform construction and use engineers.
After two months of testing, a total of more than 500 enterprises have applied for trial, combined with customer experience needs and feedback, the product has been constantly improved and polished, the commercial version and the previous beta version have a lot of new changes, in the product function and experience of new upgrades.
E-hpc and traditional supercomputer center, E-HPC is purchased according to the required scale, no scheduling, buy and use. E-hpc has a wide variety of models and fast hardware update. You can purchase the most suitable ratio of computing resources (such as ECS instance) according to application requirements. E-hpc fully reuse Ali Cloud products, greatly improving data security and high availability. Connect cloud services in an event-driven way to quickly build cloud-native applications. Second, e-HPC and self-built physical HPC cluster, E-HPC can save a huge amount of CAPEX input, including but not limited to: hardware system cost; Software license fee (multi-core, multi-node); Server room construction, power and cooling costs; Maintenance and support expenses for daily operation. E-hpc does not need to worry about hardware upgrading, keeping up with the latest hardware development. E-hpc connects to the whole series of Aliyun products with strong expansibility. Third, E-HPC and self-built cloud server cluster. E-hpc provides convenient software management services, covering all software stack functions of HPC applications, and upgrading with the HPC community (OEM/OSV/ISV/ user/developer/Open source community) versions. E-hpc provides the capacity expansion or reduction service for cluster nodes on demand, saving costs. E-hpc provides complete HPC job workflow services to match the working habits of HPC users.
The massive IDC(Internet Data Center) infrastructure built by Aliyun provides universal computing resources for the whole society. Naturally, the elastic high-performance nodes through Ali Cloud will “pop up” the HPC stack comparable to small and medium-sized supercomputing centers to meet the needs of industrial applications. This is ali Cloud E-HPC products. Based on ali Cloud’s high performance SCC Cluster (Super Computer Cluster), ECS (elastic computing service), EGS (elastic GPU service) and other computing products, Ali Cloud’s elastic high performance platform E-HPC has strong parallel computing capability and is perfect for high performance computing business demands.
E-hpc has other advantages over traditional supercomputing centers, such as: Flexibility: HIGH-PERFORMANCE computing clusters can be created and released on demand without waiting for applications. During running, the e-HPC cluster scheduling policy dynamically scales the number of nodes to provide optimal resource utilization and break through the scale limit of offline physical clusters. Fast software and hardware iteration capability: fast software and hardware iteration, software stack support the latest computing framework, chip/media keep pace with the rapid evolution of technology development, enjoy the latest technology dividend; Product connectivity: E-HPC seamlessly connects with other Aliyun products, and connects aliyun services through HPC job workflow to adapt to the working habits of supercomputer users. The job data can be quickly read and written through Ali Cloud storage and network services, and the job results can be saved to ali Cloud mass object storage and displayed through Ali Cloud desktop products, and can also be quickly synchronized to the cloud environment through Ali Cloud high-speed channel products. Regional connectivity (global deployment) : Through Aliyun global multi-regional deployment, customers can choose the place closest to them to deploy high-performance cluster, with the internationalization process of customers out of the overseas; Security: From tenant isolation and application isolation to trusted computing based on underlying hardware, Ali Cloud provides strong security guarantee.
Buying experience link: www.aliyun.com/product/ehp…