C6
Cloud

Back to overview

10:40
conference time (CEST, Berlin)
A Workflow for Combining Highly Scalable and Parallel Cloud CFD/FEA Simulations With Multi-Objective Optimisation Models
26/10/2021 10:40 conference time (CEST, Berlin)
Room: C
N. Zhukov, A. Saratov (Datadvance, RUS); A. Dammer (SimScale GmbH, DEU)
N. Zhukov, A. Saratov (Datadvance, RUS); A. Dammer (SimScale GmbH, DEU)
SimScale GmbH and Datadvance have developed a scalable cloud integrated workflow for multi-objective optimisation studies. This paper presents a case study of a quasi-static, axisymmetric thermal-structural problem in the form of a low pressure rotating turbine disk model. The purpose is to optimise the physical characteristics and geometry of the rotating disk which has 11 variable geometric dimensions and multiple constraints, whilst minimising the overall mass and keeping within structural and stress limits. An Application Programming Interface (API) is used to connect the geometric optimisation tool to SimScale’s multiphysics cloud simulation engine. pSeven Enterprise is used to build and run the optimization workflow. Each time a particular geometry and simulation parameters are sent to the API, SimScale will simulate the model and return outputs for that instance including temperature distribution, radial displacements and stresses. Data is then fed into the pSeven surrogate optimisation model. As the workflow is scalable, many hundreds of geometric scenarios can be modelled. The physics outputs for each scenario are tracked using a pareto front to converge on the local minima (the optimised solution) where the mass and radial displacement are the two key parameters. In one case, after 400 iterations, a 20% reduction in overall mass of the disk was achieved. SimScale simulated all 400 iterations on the cloud, in parallel, without the need for any local hardware, thus the speedup in simulation and analysis time is exponential compared to current approaches. The new workflow makes the procedure time and cost efficient and accessible via only a web browser, meaning, for the first time this kind of speed and power is available without the need for costly and complex hardware. pSeven Enterprise allows going beyond the solution of optimization problems in the cloud. A workflow-powered web application can be built and published to AppsHub gallery within pSeven.
Cloud computing, optimisation, API, CFD, FEA
11:00
conference time (CEST, Berlin)
Migrating Engineering Simulations to the Cloud Use Case: How Freudenberg Embraces Cloud for Infrastructure Modernization
26/10/2021 11:00 conference time (CEST, Berlin)
Room: C
W. Gentzsch (TheUberCloud, DEU); C. Weis (Freudenberg Group, DEU); L. Miroslaw (Microsoft, DEU)
W. Gentzsch (TheUberCloud, DEU); C. Weis (Freudenberg Group, DEU); L. Miroslaw (Microsoft, DEU)
Engineers have long suffered from inadequate systems and software. Industrial surveys show that about 60% of engineers feel that hardware resources are inadequate for their work. Often, they have to simplify models to fit in current hardware limits thus losing accuracy of their results. In addition, buying HPC system comes with even more challenges, such as long procurement cycles; high Total Cost of Ownership (TCO); non-scalable software; erratic system usage; fast aging of inhouse servers; and the need for trained HPC system experts. But with the rise of cloud computing, it's easier than ever to get access to powerful computing, efficient storage, and low-latency networking. More and more organizations are turning to HPC cloud, due to additional business value coming from increased flexibility and scalability from practically unlimited, on-demand computing and storage capacity that can be provisioned if needed. The productivity of the simulation team grows thanks to the ability to run more projects at the same time or running more complex simulations, without big investments. Compute-intensive tasks such as parametric sweeps, multiphysics simulations, and large-scale optimization to evaluate many design options become possible. The other benefits include the possibility to increase the accuracy of their models by removing memory and CPU limitations so engineers can use more detailed models, access heterogeneous architectures such as GPUs, InfiniBand, large memory, latest processors, etc., and, finally, reducing engineering and licensing expenses by shorter computation time. However, running complex engineering workloads in the cloud is not easy. The engineering team needs cloud-specific know-how, the integration with existing on-premises infrastructure to ensure security and compliance, hands-on experience with high performant architectures, on-boarding complex applications to the cluster, optimizing software licensing cost for the cloud, and efficiently transfer large datasets. All components of compute, storage, and visualization should be optimally integrated with on-premises infrastructure into a single system. The UberCloud Platform is addressing these challenges by providing engineers a turn-key scalable engineering platform with automatically integrated compute, storage, and networking, optimized for specific CAE software, enabling engineers to significantly reduce IT overhead and modernize their IT infrastructure. The platform provides access to the latest hardware optimized for specific applications, high-resolution engineering virtual workstations for pre-post-processing and interactive simulation, scalable HPC, and automated update process without downtime, with minimal IT overhead, and the possibility to autoscale the cluster with tailored hardware for each specific job. We have demonstrated the feasibility of this approach at Freudenberg Technology and Innovation (FTI), an engineering facility providing expertise and simulation services to the Freudenberg Group, a global technology manufacturer with production sites in Europe, Asia, Australia, North and South America, and more than 48,000 employees. FTI faced the challenges of continuously updating the existing simulation environment to hardware for best performance and providing tailored hardware to their engineers for each specific simulation workflow. As part of their digital transformation, Freudenberg decided to migrate their engineering workloads step-by-step to Azure to provide engineers their customized simulation workflows on Azure, including a cloud-based engineering workstation. By moving their workflows to the UberCloud Platform on Azure, seamless hardware update, best performance, and tailored hardware configurations for each specific simulation task have been provided. In our presentation, we will explain the technical details of the migration process such as the cloud templates, the automated pipeline for the deployment, the simulation workflow containerization technology, and the software management. We will also cover the integration aspects of the UberCloud Engineering Simulation Platform, the CAE application containers with Azure CycleCloud, NetApp storage, and visualization. Last but not least, we will present benchmark results of key CAE applications running fully in the cloud on the latest AMD architectures.
Engineering Simulations, CAE, Ansys, Cloud Computing, Digital Transformation, Cloud Migration
11:20
conference time (CEST, Berlin)
Challenges and Potentials of Cloud-native Software Architectures Towards Industrial use in the Field of Computational Engineering
26/10/2021 11:20 conference time (CEST, Berlin)
Room: C
J. Gutekunst, I. Simonsmeier (dive solutions GmbH, DEU)
J. Gutekunst, I. Simonsmeier (dive solutions GmbH, DEU)
Today, the productive implementation of computational fluid dynamics introduces many layers of complexity to the product development process e.g., large recurring hardware cost due to high performance computing demands, software maintenance routines and a strong data storage and administration strategies to manage the large, partially unstructured and diverse sets of data to archive and evaluate. In the last years, large cloud infrastructure providers further developed and promoted their services in the field of computer-aided engineering (CAE). The present work wants to give an overview of the most important technological advances in the field of cloud-native software architectures including the potential for more productive cloud-based engineering toolchains. First, the modern web- and cloud-based services are discussed from a software architecture perspective and evaluated towards the added value they potentially provide to the field of computational engineering. Here, the focus lies on popular topics like high performance computing in the cloud, centralized large data and microservice architecture, as well as how they can potentially contribute to the scalability and robustness of high demand services. Furthermore, the challenges and potential solutions in client-server communication incl. user authentication, management and transfer of large data sets, network latency and job prioritization are explained. In this context, the issue of data security is discussed and a holistic view on the data transfer from a client to a server, incl. potential vulnerabilities, is provided. Lastly, the introduction of a cloud-native infrastructure comes with the necessity of a fundamental business model refactoring. Thus, the cost structure and software-as-a-service (SaaS) business model behind an exemplary cloud-native CAE environment is presented based on existing industrial examples. This includes the overall value proposition, technological demands and general differentiation that a cloud-native CAE tool provides compared to a conventional on-premise software.
cloud,simulation,SPH,CFD,engineering
11:40
conference time (CEST, Berlin)
Enabling Fast, Multi-cloud HPC to SPDM Solution Integration
26/10/2021 11:40 conference time (CEST, Berlin)
Room: C
R. Klein (Rescale Southern Europe, FRA); M. Nicolich (ARAS Corporation, USA)
R. Klein (Rescale Southern Europe, FRA); M. Nicolich (ARAS Corporation, USA)
Reliance on simulation results is ever-increasing, resulting in larger model sizes incorporating more physics, larger meshes, higher numbers of iterations, and more sophisticated multidisciplinary simulation workflows. The complex high-fidelity virtual test scenarios are modelled using different simulation software packages and run on local high-performance computers (HPC), taking days or sometimes even weeks to return results. Simulation environments face unique challenges: fragmented software and hardware, a large simulation data set, and a complex execution process. Data is often isolated and managed using ageing, legacy technology. For example, simulation data is managed on the engineering desktop, or at best, through a shared NAS relying heavily on naming conventions. Files are shared with remote users via email or FTP. This is incredibly inefficient as it can be impossible to find which file was used for which analysis. With simulation processes being up to 20 steps long this can be hard to track with a reliance on files in folders. Simulation Data Management is a technology trend that has existed since the year 2000, aiming to build a simulation method, provide traceability and increase productivity through automation. “Despite the successes achieved with SDM, the adoption of information systems to manage simulation data by simulation engineers is still very low at 1%-2%,” according to NAFEMS. This is primarily down to the technology not being well understood, not being accessible to small teams and a relatively high cost. However, companies such as BMW Audi, Porsche, and certain divisions of Airbus have been using it extensively. In 2020 ARAS, Rescale and other Digital R&D-focused startups took part to a hackathon hosted by McKinsey & Company. Following this event, they explored the feasibility of plugging HPC job submission software into SDM so that engineers could both submit jobs from within their SDM environment and recover the results there too. This resulted in an MVP called eSteam. This presentation will detail the implementation of the eSteam project and how the provision of REST API’s across both platforms was the key to rapidly integrating the platforms in only a matter of weeks leading to significant improvements in resource and productivity, and allowing teams to collaborate exclusively within one environment.
Cloud, HPC, SPDM, REST, API, Open Engineering Platform, MVP Prototype, Collaboration
×

[TITLE]

[LISTING

[ABSTRACT]

[DATE]

[ROOM]

[KEYWORDS]