Inside HPC & AI News | High-Performance Computing & Artificial Intelligence

Today Pointwise announced the latest release of its computational fluid dynamics (CFD) meshing software featuring direct integration with overset grid assembly (OGA) software.

"The UberCloud Community brings together over 3,000 participants to pave the way for the Technical Computing Cloud. UberCloud Experiment teams explore the end-to-end process of accessing remote computing resources in HPC centers and in the Cloud to study and overcome the potential roadblocks for industry applications such as CAE, Bio and Life Sciences."

"An increasing number of GPU enabled applications are available to the HPC community. The key issues are understanding the enhanced application performance and corresponding increase in power consumption due to GPUs. In most cases these depend on the CPU to GPU ratio and the way GPUs and connected to CPUs. Latest compute node designs allow flexibility to select the number of GPUs and how they are connected CPUs. This offers users a unique opportunity to select the a suitable operating point according to their application characteristics. This talk is about studying the performance vs. power tradeoff on a few common HPC applications."

"Through the advent of cinema, video games, in-vehicle Infotainment and virtual reality, the industry has created a roadmap for the future. A true quality upgrade of the overall audience experience is on-going. HPCmusic is here to keep music and sound on par with those innovations and exceed the expectations of the audience. We introduce the Aural Computing Engine to bring Digital Audio 2.0 to life."

"Over the last five years HPC performance growth has been slowing measurably, and in this presentation several reasons for this slowdown will be analyzed. To reach usable exascale performance over the next decade, some fundamental changes will have to occur in HPC systems architecture. In particular, a transition from a compute centric to a data movement centric point of view needs to be considered. Alternatives including quantum and neuromorphic computing have also been considered. The prospects of these technologies for post-Moore’s Law supercomputing will be explored."

This week ThinkParQ launched BeeGFS on Demand parallel storage offering. Dubbed BeeOND, the new offering allows users with HPC clusters to generate job-dedicated parallel filesystem instances on demand.

"As cloud resources increase, users should be cognizant of the HPC platforms available in the cloud. Scalable and customizable resources are readily available, however, security, ease of use, and full integration of simulation software are important factors to consider when choosing to run analyses remotely. Cloud resources designed for engineering and science CAE simulations allow users to leverage high performance computing to fully explore the design space and develop innovative products. In this presentation, Joris will discuss the cloud HPC market trends, CAE simulations using cloud HPC, cloud software integration for CAE use, security around HPC cloud environments, data management and collaboration."

"Alongside the increasingly high demands of streaming bandwidth in HPC storage solutions, there is a growing need for higher levels of metadata performance for various applications and workloads. The Lustre parallel file system provides a distributed namespace, divided across multiple metadata servers, that allows the metadata throughput to scale with increasing servers. This presentation addresses meeting the increasing requirements for high performance metadata in Lustre environments with the ultimate aim of reducing the time to results and improving overall efficiency."

Å koda Auto, was in need of scalable, high performance computing systems to perform complex product performance and safety analysis. In this case study we learn how they used HPC to rapidly develop more innovative vehicles with excellent value-to-power ratios.

"Following up on the “Streamlining Research Computing Infrastructure: A Small School’s Perspective”, this talk will discuss how a researcher’s priority in a shared HPC cluster is computed based on her/his usage pattern as well as productivity. Coupled with transparent functionality and reporting scheme at nearly every level, this merit based priority scheme has enabled consistently high usage (85+%) throughout 20 months of operation and has helped researchers at Michigan Technological University to produce 30+ publications from approximately 30 different projects."

In this video, former SC Conference Student Volunteer Christine Harvey describes how volunteering led to her current career at Mitre Corporation.

"As HPC systems and solutions become more focused on specific workloads, the need to address system level power, cooling, and power management while maintaining usability remains. This presentation will review recent developments within several of SGI’s platforms. These developments enable socially responsible computing and better use of power and cooling."

"PayPal is a world-class leader in online and offline payment innovations. This tradition continues by combining enterprise and HPC architectures to deliver integrated solutions to large scale compute and analytics problems. This presentation will discuss the need for HPC, the growth and application of HPC and discuss innovative work that will radically change the future of real time analytics."

The UK Met Office is rolling out the Allinea DDT debugger on its newest supercomputers. The debugging tool will be used to help the scientists and developers to develop their widely used forecasting software for the new Cray XC40 systems.