AMD has submitted this post.

As manufacturers of all sizes struggle with cost and competitive pressures, and as products become smarter, more complex and highly customized, the use of computer‑aided engineering (CAE) is growing.

Using CAE, engineers can design and test ideas for new products without having to physically build many expensive prototypes. This helps manufacturers lower costs, enhance productivity, improve quality and reduce time to market by primarily focusing on designs with the best potential for market success. CAE also helps drive innovation and enhance collaboration throughout the supply chain, while mitigating risks and costs associated with potential product failure and associated litigation.

Of the CAE disciplines, multiphysics is among the most demanding because it combines several CAE applications such as structural analysis, fluid mechanics, mechanical dynamics and electromagnetics. These comprehensive, high‑fidelity simulations can help accurately predict how complex products behave in real-world environments. Iterative design exploration studies are also being extensively used to simulate, design and optimize complex systems. Multiphysics design optimization studies need very detailed geometric models and large meshes over thousands of operating scenarios. This puts enormous stress on the high-performance computing (HPC) infrastructure required to power CAE workloads.

Customer Requirements for HPC Solutions

Manufacturers of all sizes need a highly reliable, secure HPC environment that scales and performs at extremely high levels to deliver fast time to results on large, complex simulation models. The solution should also foster collaboration throughout the supply chain and reduce complexity and total cost of ownership (TCO), including capital and operational (facilities, labor, and CAE software license) costs.

As high-resolution models become larger and more complex, memory and available cache have emerged as important requirements for data centers running CAE applications. Simulation models increasingly consist of millions of elements or cells, and the more of a model that can fit in CPU memory and cache, the faster a simulation can be performed. In addition, the use of digital twins—virtual representations of physical objects that mimic their real-world counterparts—is expected to double in the next five years. Similarly, artificial intelligence (AI) and machine learning techniques are expected to triple in applications ranging from manufacturing processes to product quality to predictive maintenance. All of these requirements are driving the need for increased simulation capacity.

To meet these evolving requirements, manufacturers need:

• High‑frequency processors delivering high per‑core throughput and high core count processors to complete more jobs faster to maximize their investment in expensive ISV applications.
• Large memory capacity, high memory bandwidth and high ratios of cache per core to further improve compute performance.
• High I/O performance to improve storage performance.
• Low network latency and high network bandwidth to ensure better scaling.
• Reliability, availability and serviceability (RAS) to minimize downtime costs.

Besides raw performance, energy efficiency is also an important consideration. As engineering IT data center managers running CAE simulations seek to reduce their carbon footprint for sustainability, they need servers that deliver maximum throughput per watt to minimize power and cooling requirements. They may also need denser, more energy-efficient designs that deliver superior performance while minimizing cooling and data center space requirements.

As the market leader in HPC systems with a 33.4 percent market share, Hewlett Packard Enterprise delivers one of the industry’s most comprehensive CAE solutions across compute, interconnect, software, storage and services delivered on‑premises, hybrid or as a service. By teaming up with AMD, HPE delivers exceptional performance, flexibility, and choice on a range of CAE applications.

Visit AMD to learn more about their HPC solutions.