• banner pc

    Manufacturing CAE Simulation Solution

    Implement HPC for engineering simulation.


  • Overview
  • Benefits
  • Architecture
  • Products

Optimize Product Design

It's simple: Computer-Aided Engineering (CAE) simulation significantly optimizes product design. Powered by High-Performance Computing (HPC), the solution speeds up computing and development and overcomes every challenge potentially in the way.

  • Difficult Collaboration

    Processing tools for different simulation phases use different storage protocols. Data therefore needs to be migrated multiple times, dragging down efficiency.
  • High Concurrency

    Unlike traditional transaction-based Information Technology (IT) systems, CAE simulation requires high-concurrency access and computing.
  • Complex Applications

    With many different simulation workloads spread across a variety of disciplines, different systems are needed. That potentially makes simulation inefficient and difficult to manage.


Precise Defense

Multi-Protocol Interworking

• During simulation, pre-processing files are read directly, doing away with the need for data migration.

Open Edge Computing

Hardware and Software that Supports High Concurrency

• From CPUs and NPUs to BMCs.
• Kunpeng processors, Ascend AI chips, network interface cards, intelligent management chips, and SSD control chips.
• Full-stack software, such as parallel and math libraries.
• MPIs, scheduling management software, compilers, and more, optimize manufacturing applications.

Management Platform

Complex Workloads

• OceanStor Pacific 9540 storage systems support high-concurrency and high-OPS operations.
• System read-write performance exceeds 30,000 OPS, with a single thread providing bandwidth of 400 Mbit/s, improving simulation performance.


Huawei provides Information and Communications Technology (ICT) infrastructure including data center computing, storage, and switches for fluid simulation, structure, electromagnetism, and other domains — addressing every potential challenge. The solution addresses a series of core factors, including the scheduling system, communications library, compiler, cluster management, and services.

en jiagoutu