CUHK Advances Scientific Research with Next-Gen HPC Powered by Huawei Data Storage

Bridging the Gap Between Research Ambitions and Data Infrastructure

Simply put, universities require the next generation of integrated HPC platforms to meet the diverse needs of different research fields while continuing to evolve, providing reliable, long-term data support for scientific research. And striking the right balance between high-performance data access and mass research data storage—as well as building a storage platform that meets current needs while supporting future growth—have become the key challenges to face.

In this environment, the Chinese University of Hong Kong (CUHK) was looking to advance its research infrastructure and wanted to build a next-generation integrated HPC platform that could provide computing and data support for the diverse needs of its multiple faculties.

All-Flash and High-Capacity Storage Meet Diverse Research Needs

In HPC scenarios, different research tasks have different requirements for storage resources. Some research tasks are data-intensive and require frequent access to datasets, demanding high data read/write efficiency to help compute resources operate efficiently. Meanwhile, the significant amounts of data generated throughout the research process must also be stored efficiently—and cost-effectively. In addition, storage systems must be able to accommodate growing capacity demands while optimizing the physical footprint and overall resource utilization.

To meet these dual demands—performance and capacity—storage infrastructure must break free from the limitations of a single product form and allocate storage resources based on workload characteristics. With this in mind, the CUHK chose Huawei OceanStor Pacific 9920 and OceanStor Pacific 9550 for its HPC platform. Featuring all-flash storage and high-density HDD storage, respectively, the storage systems balance high read/write performance and cost-effective, expandable high-capacity, establishing a solid data foundation for multidisciplinary research.

High-Performance Access Accelerates Scientific Computing with Efficient Data Support

In scientific computing, data access efficiency is a key factor influencing application performance. For data-intensive research tasks, the storage system must have efficient read and write capabilities and deliver data efficiently to help compute resources operate efficiently.

With all-flash architecture, Huawei OceanStor Pacific 9920 is designed for high-performance data access, offering robust storage support for scientific computing and data processing. Integrating all-flash storage into the CUHK's HPC platform provides storage resources that are optimized for demanding access workloads across a range of research applications. And, aligning storage infrastructure more closely with the needs of scientific applications provides a stronger technical foundation for research.

High Density and High Capacity Enable Efficient Storage of Mass Scientific Research Data

Scientific research data is vital to sustained university research, and as research advances, experimental data, simulation outputs, and other research materials keep growing, so the storage system must preserve existing data while supporting future data expansion.

Huawei OceanStor Pacific 9550 uses a high-density, large-capacity HDD design to deliver massive data storage capacity in a compact chassis—a cost-effective storage choice for mass scientific research data. As the scale of research data grows, high-density storage actually improves capacity utilization per unit space, providing a solid foundation for the CUHK's HPC platform.

This combination of OceanStor Pacific 9920 all-flash storage and OceanStor Pacific 9550 HDD-based storage balances high-performance data access and high-density, large-capacity storage, enabling the HPC platform to access storage resources based on research needs.

Scale-Out Architecture Supports Continuous Platform Evolution

Of course, university research is always evolving and advancing, so an integrated HPC platform must also adapt to changing needs, with storage infrastructure that fosters future research.

Unlike an architecture with fixed capacity, Huawei OceanStor Pacific's scale-out architecture is future proofed, supporting on-demand storage expansion and enabling flexible resource planning for future platform development.

As the platform serves multiple faculties and diverse research areas, its scale-out storage solution provides high performance and large capacity to meet present data access and capacity demands, and form a sustainable infrastructure for future research data growth and application evolution.

Innovative Storage Lays the Foundation of Scientific Research Development

The coordinated development of computing power and storage power is critical for innovation in scientific research. While an HPC platform delivers computing capabilities for scientific research tasks, efficient and flexible storage infrastructure guarantees the access, processing, and accumulation of data. As university research shifts towards multidisciplinary, data-intensive tasks, a storage platform that adapts to diversified needs and is capable of evolving will become an essential part of scientific research infrastructure. And Huawei OceanStor Pacific 9920 and OceanStor Pacific 9550 provide strong storage support for the CUHK's integrated HPC platform, combining all-flash storage and high-density, large-capacity arrays.

Huawei Data Storage will deepen its engagement in university research and HPC, using innovative storage technologies to upgrade research infrastructure, consolidate data foundations, and support scientific exploration and innovation.