High performance computing in engineering simulation

Designing and testing products in the virtual world is faster and less expensive than making physical prototypes, but it depends on the right level of high performance computing (HPC) as well as software.

Here our expert roundtable panel – hosted by specialist HPC integrator OCF – discusses current thinking on today’s most important engineering simulation issues.

As consumers demand ever more capability for their investment, technology disruption is rapidly becoming a fixture in the product design cycle. This disruption inevitably leads to an increase in demand for more and more product and system protypes, all delivered on time and on budget. For manufacturers under pressure to meet this demand, the need to maintain tight control over development costs for cutting-edge applications means that engineering simulation on HPC platforms is becoming more widespread than ever. Once the exclusive province of big ticket applications such as defence, aerospace and automotive, over the past few decades product design simulation has trickled down into smaller scale applications, to the point where that flow has become one of the features of product development at a more everyday level such as components for electronic gadgets.


Wim Slagter, HPC and Cloud Alliances Director at Ansys

Describing traditional engineering applications in which HPC-based simulation is finding increased importance, HPC and Cloud Alliances Director at Ansys, Wim Slagter sets the scene. “Technology disruption is making manufacturer’s lives even tougher. One area is electrification in the automotive arena. Consumers are clearly expecting new electric vehicles that match the performance of petrol-powered cars: they must have similar costs, a range of more than 300 miles on a single charge and also the ability to restore battery life in a matter of minutes. That’s a big challenge.” He goes on to say that another important area to consider is the implementation of 5G antenna infrastructure, “which needs to be designed with low cost and that can work with existing infrastructure. That’s not trivial.” The same conditions hold for the proliferation of smart connected products that are providing a gateway to the Internet of Things (IoT), autonomous vehicles, robotics and digital twinning.


Bastiaan van Brussel, Global Lead for HPC Alliances at Dell Technologies

Global Lead for HPC Alliances at Dell Technologies, Bastiaan van Brussel views the same question from a commercial competitive angle, saying: “from the perspective of a global economy in which customers are continually struggling to bring their product to market at pace – trying to keep up with their competition – there are huge pressures on manufacturers to deliver on that challenge. It’s a rat race out there. We need to understand that and provide our customers and partners with everything they need in terms of technical solutions to these challenges.”


Andrew Dean, Sales Director at OCF

This increased requirement for agility is never more apparent than in additive and subtractive manufacturing, areas in which, “we are creating products today more complex that we could have ever imagined,” says Andrew Dean, sales director at OCF, Sheffield-based provider of HPC, storage, cloud and AI solutions. What we are finding today, says Dean, is that the traditional design and manufacture methodologies, as well as the once leisurely timeframes, are simply no longer the way things get done. “This is one of the most important areas where we find engineering simulation fits in: effectively helping engineers to take advantage of the latest manufacturing techniques.” Dean is happy to admit that computer-based simulation has been well-established for decades, but what we are now seeing is more and more smaller organisations and engineering teams realising its potential for the first time and being able to take advantage of, “getting more for less. Compared with five years ago, compute power is just going up and up. I think that there’s a perception generally that engineering simulation is a large-scale method of offering projected results of what can happen in the real world. But it is now becoming more approachable, and it is being taken up in less grand challenges.”


Van Brussel notes that it’s not just a question of potential, but affordability too, as HPC these days comes more within the financial reach of a wider range of end-users across industrial sectors and organisational scale. Dean agrees, adding that while designing a product and then testing it in the virtual world before actually building a physical prototype has the key benefit of being fast, it also brings with it lower costs. “And so engineering simulation makes sense from a financial point of view,” says Dean. Slagter comments that it is also noticeable how HPC hardware and software systems have evolved to become ‘more powerful and accessible’ over the past decade, leading to them becoming, “more readily available for faster design iterations on the engineering platforms of small and mid-size companies as well.” But, he warns, for all the progress being made in this area, “misconceptions about HPC still exist.”

One of main misconceptions, says Slagter, is that “a lot of people tend to think that HPC is complex.” He goes on to say that there’s a good reason for this, because “in principle it does not always look simple, because it is about sizing, building, integrating, provisioning and supporting. You need specialised IT skills expertise, and these skills are often unavailable to users of engineering simulation software. But clients need this support environment to set up a viable system, and that is why you definitely need to work with partners.” There is a perception that there is an inherent complexity in such a system, “but that can be mitigated against,” explains Dean, who goes on to say that the key here is, “to work with the right partners and with the right technologies – there are software products out there that make adopting these technologies much easier. Also, training and support will help the user to integrate these solutions as easy as possible.” Van Brussel sums up the situation by saying that, “whether the situation is actually complex, or it is just perceived to be complex, the way to master that complexity is through partnerships – through the right tools, technology and people. The degree in which companies are able to master complexity – in terms of the product they engineer and the tools they leverage for this objective – has a strong impact on the level of innovation they can achieve and the frequency in which they are able to innovate. Ultimately, companies that innovate more and innovate faster compared to others, will have a competitive edge.”

If manufacturers get HPC tightly integrated with engineering simulation, concludes Dean, “it should become just another tool for engineers. Selecting HPC systems should be as straightforward as choosing a printer. That’s how tightly integrated we want these systems to become, so that they become easier and easier to use. That’s a key thing that HPC systems give you: high utilisation and scale.” He further explains that while the majority of today’s engineering simulation is performed on discrete workstations by engineers working in isolation, what an HPC system can do is bring these single sources together, “into one big pool in the middle, so that everyone shares, which means that each individual engineer now has access to a much bigger resource. This in turn means that they can get results back quicker because they can run the same simulation over many more machines, or they can run more complex challenges and expect to get the results back in the same time. Basically, HPC enables engineers to do things that they previously weren’t able to do.”

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles