Machine Learning for Accelerated Aero-Thermal Design in the Age...

pareto front

Enhanced Topology Optimization with Multi-Objective Continuous Adjoint

March 28, 2020 Comments (0) Views: 1052 Demonstration, Video, Webinar

Cloud Computing with HELYX and Penguin Computing On-Demand

Companies using simulation in their design process, often utilize a spectrum of computing resources to meet budgetary and uptime requirements, manage security risks, and remain flexible in order to use only as much compute time as needed.  The types of computing paradigms include running on local workstations; local high performance computing (HPC); remote HPC; and third-party HPC providers.  Each option possesses different costs (CapEx vs. OpEx) and benefits, where it’s up to companies to balance budget and benefits to achieve an optimal portfolio of computing resources. 

We’ve seen where HELYX GUI’s client-server technology enables users to connect to different computing resource types (local, remote, and third-party).  You may have also seen where HELYX can directly access resources on EPCC through HELYX On-Demand services.  This approach has proven to be a powerful tool for users, facilitating the connection to computing capability where it is available.  We’ve worked to offer even more options to HELYX and ELEMENTS customers, by providing access to HELYX and ELEMENTS through Penguin Computing On-Demand (POD).

POD offers bare metal HPC (as opposed to virtualized compute), enabling greater parallel scaling of larger simulations.  In addition to bare metal resources, POD offers ENGYS customers a way to:

  • Work seamlessly within a browser-based cloud workstation provided via POD
  • Access thousands of cores to enable fast and scalable compute for large simulations
  • Quickly deploy HELYX to their teams and start working with the powerful meshing, simulation, post-processing, and advanced optimization tools within HELYX and HELYX Add-On modules.
  • Submit simulations directly (meshing, setup, solving, and post-processing ) from HELYX GUI through a specially configured client-server run-mode. This allows users to switch between queues and submit jobs through the click of a button.
  • Consolidate software and compute vendors, using an ENGYS maintained account within POD, equipped with the latest ELEMENTS and HELYX binaries.

Overall, customers need access to a variety of computing resources and through working with POD, thousands of bare-metal computing cores are easily within reach. 

Take Home Messages

Increase the availability of high performance computing to your organization and scale as needed with HELYX and ELEMENTS on POD.  Work seamlessly within a browser based workstation or work directly within a shell to submit jobs to thousands of bare-metal cores.  Contact ENGYS for more information how to use HELYX on POD and how to trial our software before buying.  

Learn More by Watching the Webinar

Learn how HELYX customers can easily access on-demand, high-performance bare-metal computing resources from Penguin Computing On-Demand (POD). While on the POD platform, users can mesh, setup, solve, and post-process with HELYX on the Scyld Cloud Workstations and submit directly to a configured HPC queue.  Joined by Will Cottay, Penguin Computing Director of Cloud Services, attendees will learn more about POD and obtain a sense of how to effectively use HELYX on the POD platform. The webinar agenda follows:

  • Motivation (Starting at 5:00 )
  • POD Overview and Demonstration (Starting at 8:00 )
  • HELYX on POD (Starting at 18:30 )
  • Take Home Messages (Starting at 43:57 )