When using simulations (mechanical or computer based), it becomes possible to make statements about the potential behavior of a subject, without a direct interaction with the physical entities. Prediction of the potential behavior of subjects under changed conditions is thus one of the most important roles of simulations. As just one obvious example, performing initial crash-tests with a computer-implemented model of a car instead of a real one will certainly save money even in the short run and, due to the ability to test many different scenarios, might even make the car (far) safer than it would be using time-consuming physical tests alone.
Parametric optimization techniques make it possible to automatically determine parameters of the chosen model that maximize or minimize specific properties. Here simulations fill the role of a solver, helping to determine one or more figures of merit connected to a given parameter set and model. Thus simulations pave the way to new designs that, in all likelihood, would not have been found using human expertise alone, due to the usually high dimensionality of the input space of the model.
As simulations may also require significant amounts of compute time, many simulation types are closely related to HPC (although most computing devices, from a smart phone to the most sophisticated super computer, are capable of running at least some type of simulations). And, as modern computing devices also become increasingly more powerful, models and thus simulations may become more detailed over time, giving them a better ability for valid predictions and thus raising their significance for society.
It is important to note that simulations also affect many aspects of daily life, e.g. when estimating the spread of epidemics or simulations of pollutions in atmosphere.
In general, simulations, often in conjunction with powerful HPC devices, may provide us with tremendous opportunities that we should use. Some simulations even represent the only tool we have to make at least a rough forecast of what is to come. Not using them would be akin to driving blind-folded on a highway.
Stuttgart, 13. November 2018
The High-Performance Computing Center of the University of Stuttgart (HLRS) and Hewlett Packard Enterprise (HPE), the number one market leader in high-performance computing (HPC), today announced a joint collaboration to build and deliver for HLRS a next-generation supercomputer, 3.5 times faster than its current system. The upcoming system, which HLRS has named Hawk, will be the world’s fastest supercomputer for industrial production, powering computational engineering and research across science and industrial fields to advance applications in energy, climate, mobility, and health. Hawk, based on HPE’s next-generation high-performance computing (HPC) platform running a next generation AMD EPYC™ processor code named Rome, will have a theoretical peak performance of 24 petaFLOPs, and consist of a 5,000-node cluster.