HIGH PERFORMANCE COMPUTING SPONSORED CONTENT
The increasing complexity of simulation workloads is driving a need for HPC resources in computer-aided engineering, finds Robert Roe
Engineering drives new simulation tools
As increasingly competitive markets tighten product development cycles, the need for innovation rises. Many organisations also want to reduce expenditure and pool resources to deliver a more efficient computing infrastructure, which is increasing the use of cloud and threatening the traditional general- purpose HPC cluster. These factors are driving simulation engineers across several industries – from traditional HPC users such as aerospace and automotive, to more recent adopters, including construction, robotics, autonomous systems and healthcare – to turn to HPC simulation. Organisations need to increase capacity for simulation, while delivering a better-optimised platform that can be part of a strategy to reduce costs.
The need for HPC The rising use of digital twins to shorten product development and reduce the costs of validation and testing is also increasing the demand for computational simulation resources. Today, engineers need high-performance computing and large scale simulation so they can deploy new materials, refine their understanding of composites or to better understand physical phenomena such as combustion or turbulence, with more fidelity than was previously possible. This challenge is compounded by the need to push the boundaries of simulation, across CFD, materials physics or interactions between components and systems, which also drives the need for more complex and larger simulations. The challenges of developing innovative products in shorter timespans increases the pressure on organisations to deliver more from each development cycle. All these use cases are driving a need
for more complex and higher fidelity simulation across the majority of verticals engaged in computer-aided engineering (CAE) simulation. While traditionally HPC may have been employed by oil and gas,
14 Scientific Computing World Autumn 2020
automotive and aerospace users, today it proliferates across all verticals in CAE). A Grand View Research report in
February found that ‘the global CAE market size was valued at $7.3bn in 2019 and is expected to register a CAGR of 9.3 per cent over the forecast period’ (2020- 2027).1 It also notes that the industry is shifting
from on-premise to cloud computing in order to reduce costs of hardware provision and software licencing. However, these same market drivers are
also influencing the development of pre- configured simulation platforms, which are effectively HPC appliances that are specifically designed and optimised to run HPC CAE simulations with common ISV applications. Intel Select Solution2 portfolio are
pre-configured and benchmarked computing solutions designed for a variety
of computing applications, including HPC. The modelling and simulation variant, delivered through a partnership programme with several HPC integrators, is designed for CAE simulation users that want to accelerate infrastructure deployment. The solutions are designed to overcome challenges with hardware selection and software performance to ensure optimised engineering simulations. This enables organisations to scale quickly with a workload optimised solution that contains the latest HPC, a high bandwidth fabric and fast storage.
Increasing efficiency HPC provides performance but this comes at a cost of infrastructure and expertise to get the most out of a cluster. Without the correct optimisation and staff knowledge, an organisation will
@scwmagazine |
www.scientific-computing.com
Gorodenkoff/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38
