disc-hanging

As we noted in an earlier blog post about our Cloning feature (http://blog.rescale.com/cloning-on-rescale/), engineers and scientists spend significant time managing their simulation data. This is a problem because it interrupts their focus on product development and hinders progress in creating the best conceptual, preliminary, and detailed designs possible.

At its core, building great products requires a tremendous amount of iterations. Today, a major trend is that this iteration process is happening much earlier in the product development cycle. This is largely attributed to the ever-increasing adoption of simulation and other computer-aided analyses.

As simulations get more complex, input and output files are getting larger (at Rescale, we have dealt with customer output files larger than 100GB). In addition, there are now more files needed per simulation. Even if we isolate a single engineering project, it involves multiple iterations of analyses. A single run contains at least one input file, along with many custom scripts and add-on files. On top of this, different versions of the input files exist separately in order to track various changes. Finally, multiple analyses result in multiple output files. Add this to the need to share files between colleagues, or even the need to make copies of these files to run derivative analyses, and it can get complicated quickly. So much data exists, and a great deal of it is redundant or unnecessary. All this equates to engineers and scientists wasting precious time looking through this ever-growing mountain of data.

At Rescale, we recognize the seriousness of this problem, and have designed solutions to address the issue. At the base level, we provision low-cost cloud storage to our customers. Beyond that, our Clone and Share features eliminate the need for engineers and scientists to (a) find the necessary files, (b) associate those files with the correct analysis, (c) transfer those files either to another folder or to a colleague’s folder, and (d) ensure that  transferred files are the correct version. This saves time and effort – resulting in more time spent solving real engineering and scientific problems.

We continue to work towards finding the best solutions to this pressing problem. To learn more about Rescale, please visit, www.rescale.com.  To learn more about how you can reduce the time you spend managing your data, please contact us at info@rescale.com.

This article was written by Rescale.

reup-winners

San Francisco, CA – Monday, September 9th marked the submission deadline for the first annual Rescale University Program (ReUP) Simulation Research Grant competition. The competition was created to enable engineering and science researchers at universities around the world to run their simulations with ease using high-performance cloud computing.

Applicants were required to submit a short proposal outlining the research issue they were looking to solve, as well as how they planned to use Rescale’s flexible and secure web-based platform.

Rescale was impressed with the high quality of submissions, and we are proud to present the winners of the 2013 ReUP Simulation Research Grant competition.

Award Picture Name Institution Research Topic
1st Place Marsden, Alison, Ph.D. Alison Marsden, Ph.D. University of California San Diego “Blood Flow Simulations to Advance Pediatric Heart Surgery.” Dr. Marsden will utilize an in-house CFD software code on Rescale to model blood flow simulation in order to optimize pediatric ventricular assist devices.
1st Place Kao, EvanRayz, Vitaliy, Ph.D.Saloner, David, Ph.D. Evan Kao / Vitaliy Rayz, Ph.D. / David Saloner, Ph.D. University of California San Francisco “Numerical Simulations of the Flow in Arteriovenous Fistula Created for Dialysis Access.” Kao, Rayz, and Saloner’s proposed research includes performing CFD and FEA simulations with ANSYS software to model the fluid-solid interface and wall stresses for arteriovenous fistuals (AVF).
3rd Place Abdelal, Gasser, Ph.D.Robotham, Anthony, Ph.D. Gasser Abdelal, Ph.D. / Anthony Robotham, Ph.D. Queen’s University Belfast / Aukland University of Technology “Numerical Simulation of Lightning Strike Effects on Composite Structures.” Doctors Robotham and Abdelal will be using the Abaqus FEA software on Rescale’s platform to simulate lightning strikes, aiding to the study of delamination in composite material properties.
Honorable Mention Poh Ai Ling, Amy, Ph.D. Amy Poh Ai Ling, Ph.D. California Institute of Technology Numerical modeling of city energy consumption
Honorable Mention photo_unavilable Selin Arslan, Ph.D. Lawrence Technological University “Computational Analysis of Evaporation in Tailored Microchannel Evaporators.”
Honorable Mention Mukherjee, Sumanta Sumanta Mukherjee Indian Institute of Technology FEA simulations of orthopedic implants

All winners will be awarded Rescale credit to run simulations on Rescale’s platform. Our two first place winners will receive $2000 Rescale credit each, and our third place winner will receive  $500 Rescale credit. In addition, we are also pleased to present three Honorable Mention Awards in the amount of $100 Rescale credits each. All other applicants will receive $50 Rescale credit.

Researchers have until January 17th to conduct their simulations and submit a case study regarding their analyses.

For more information about Rescale, please visit: http://www.rescale.com/.

This article was written by Ilea Graedel.

pieces

San Francisco, CA – Rescale and Convergent Science are pleased to announce a strategic and technology partnership that will provide users of CONVERGE™ CFD software the ability to run their simulations in the cloud with pay-per-use pricing using Rescale’s end-to-end, web-based simulation platform. In conjunction with the partnership, Convergent Science is also introducing CONVERGE™ in the cloud pricing, a highly flexible option designed to offer customers the most scalability for the best possible value.”

“Many of our customers are running increasingly complex CFD analyses using CONVERGE™. As a result, they are demanding highly scalable computing and licensing solutions. Rescale is the perfect partner – they provide a turnkey solution geared towards engineers and scientists. We have been working with Rescale for some time now on various customer projects, and we are very excited to make this partnership official and announce CONVERGE™ in the cloud pricing,” noted Rob Kaczmarek, Director of Sales and Marketing for Convergent Science.

Rescale provides a comprehensive suite of features – on-demand access to high-performance computing, full integration with simulation tools such as CONVERGE™, and an intuitive user interface – all delivered through any web browser. In addition, Rescale also adheres to the highest industry standards for security at every level of the Rescale experience; engineers can transfer, manage and store data with the utmost confidence. Interested customers should approach Rescale directly for more information on running projects with CONVERGE™ in the cloud pricing.

“CONVERGE™ is a powerful CFD software tool that eliminates the user time needed to generate a mesh. We are thrilled to partner with them to help customers run CFD analyses on-demand in the cloud. No more computing or licensing bottlenecks means that CONVERGE users have an unprecedented amount of freedom when it comes to running CFD analyses,” said Rescale CEO Joris Poort.

For more information about Rescale, Convergent Science, and CONVERGE™ CFD:

Rescale: http://www.rescale.com/
Convergent Science: http://www.convergecfd.com/
CONVERGE™ CFD Software: http://www.convergecfd.com/products/converge/

About Rescale:
Rescale provides a secure, pay-per-use, web-based platform that helps engineers and scientists build, compute, and analyze large simulations on demand. Incorporated in 2011, Rescale is located in San Francisco, CA and works with customers in the aerospace, automotive, oil & gas, and life sciences industries.

About Convergent Science:
Convergent Science Inc. and its software CONVERGE™ are revolutionizing the Computational Fluid Dynamics (CFD) industry. CONVERGE™, is a CFD software package that is changing how fluid dynamics modeling is being used by engineers, researchers and designers. With CONVERGE™, the traditional CFD bottlenecks of grid generation has been completely removed from the CFD workflow, allowing users to spend more time analyzing their simulations.

This article was written by Rescale.

Recently, we completed a Rescale project with an aerodynamicist at Bombardier Aerospace. With more than 35,500 employees and a leadership position in global markets, Bombardier Aerospace designs, manufactures and supports innovative aviation products for the business, commercial, specialized, and amphibious aircraft markets.

This project was conducted to demonstrate that industrial-scale simulation in the cloud is not only viable, but also represents the best choice for both large and small companies regarding most engineering and scientific applications. Rescale worked on this project in conjunction with the HPC Experiment organized by Ubercloud. We used open-source data and open-source simulation tools exclusively.

This project involved setup and simulation of unsteady airflow around aircraft landing gear. Rescale provided the browser-based simulation platform used for set up and submission of the job, as well as the computing capacity demanded by customers, on the order of thousands of computing cores. The software code of choice was Gerris, and the required model files were sourced from the American Institute of Aeronautics and Astronautics (AIAA).

As is typical for Rescale projects, the customer first uploaded the model files securely to Rescale using our proprietary file transfer manager. Rescale’s developers then helped set up the first trial run, including selecting software, hardware, and job settings – a process taking less than 5 minutes for most jobs.

vorticity-plot

Visualization of vorticity around aircraft landing gear 25% through simulation illustrating the Gerris built-in grid refinement

Rescale is also capable of providing real-time visualization of analysis results. The vorticity plot shown above was displayed directly on the user’s web browser through the Rescale platform.

Thanks to Rescale, the engineer was able to procure all the computing capacity needed for this job in seconds, and simultaneously run analyses over a period of days. The final analysis took more than 10 days (240 hours) to converge, used 16 cores (as requested by the customer), and required 200GB of RAM memory. To maximize Gerris’s performance, the entire simulation ran using “in-core” memory, requiring an industry-leading 15GB per compute core. Upon completion of the analysis, the engineer downloaded 60GB of results file data for further post-processing on his local desktop.

When asked about his overall experience using Rescale, the customer said “Rescale was very responsive when it came to installing and getting Gerris running on their system. The web interface is intuitive and offers enough commodities to monitor complex CFD jobs on the fly (via log file tailing and flow visualization snapshots displayed by the software). From my perspective, this is a very effective solution for engineers doing large CFD, FEA, and other such analyses.”

We believe the main benefit to end users is Rescale’s ability to provide vastly different aspects of running successful industrial-grade simulations easily through any web browser, including (a) access to simulation software codes, (b) high-performance computing (HPC) hardware, and (c) actual engineering and CAE expertise. The Rescale team comprises engineers and computer scientists with deep domain expertise, which we leverage to meet all customer demands promptly and effectively.

To learn more about doing your own projects with Rescale, contact us at info@rescale.com.

This article was written by Mulyanto Poort.