Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


NVIDIA Sets Six Records in AI Performance

This week NVIDIA announced that the company has broken a total of six performance records on a broad set of AI benchmarks. As a full suite, the benchmarks cover a variety of workloads and infrastructure scale – ranging from 16 GPUs on one node to up to 640 GPUs across 80 nodes. “Backed by Google, Intel, Baidu, NVIDIA and dozens more technology leaders, the new MLPerf benchmark suite measures a wide range of deep learning workloads. Aiming to serve as the industry’s first objective AI benchmark suite, it covers such areas as computer vision, language translation, personalized recommendations and reinforcement learning tasks.”

Video: Lenovo Leads the TOP500 with Innovative HPC Cooling Technologies

In this video from SC18 in Dallas, Scott Tease from Lenovo describes how the company is leading the TOP500 with innovative HPC cooling technologies. “At #8 on the TOP500, the Lenovo-built, hot-water cooled SuperMUC system at the LRZ in Germany is one of the most power efficient supercomputers on the planet. With more than 241,000 cores and a combined peak performance of the two installation phases of more than 6.8 Petaflops.”

Red Hat Steps Up with HPC Software Solutions at SC18

In this video from SC18 in Dallas, Yan Fisher and Dan McGuan from Red Hat describe the company’s powerful software solutions for HPC and Ai workloads. “All supercomputers on the coveted Top500 list run on Linux, a scalable operating system that has matured over the years to run some of the most critical workloads and in many cases has displaced proprietary operating systems in the process. For the past two decades, Red Hat Enterprise Linux has served as the foundation for building software stacks for many supercomputers. We are looking to continue this trend with the next generation of systems that seek to break the exascale threshold.”

Ingram Micro Artizen Soutions for HPC

In this video from the GPU Technology Conference, Rick Young from Ingram Micro describes the company’s Artizen HPC solutions. “Available now to channel partners in the U.S., the distributor’s new and exclusive line of Artizen High Performance Computing (HPC) offerings include turnkey high performance servers, ultimate workstations, and customizable supercomputing clusters, as well as computing integration and software installation services.”

Managing the GPUs of Your Cluster in a Flexible Way with rCUDA

In this talk, we introduce the rCUDA remote GPU virtualization framework, which has been shown to be the only one that supports the most recent CUDA versions, in addition to leverage the InfiniBand interconnect for the sake of performance. Furthermore, we also present the last developments within this framework, related with the use of low-power processors, enhanced job schedulers, and virtual machine environments.”

Troy Baer from NICS Wins Lifetime Achievement Adaptie Award

“Congratulations go out to Troy Baer, HPC system administrator at the National Institute for Computational Sciences (NICS), University of Tennessee. Troy Baer’s contributions in scheduling and resource management using Moab have helped Kraken—NICS’ flagship computing resource and the first academic computer to break the petaflop barrier—achieve outstanding 90-95% utilization rates since 2010. Baer’s administrative capabilities enable researchers in numerous scientific arenas, from climate to materials science to astrophysics, to achieve breakthroughs not yet possible on other resources. In November 2012, Baer helped NICS’ Beacon system secure a No. 1 ranking on the Green500 list of energy-efficient supercomputers.”

DK Panda Presents: Programming Models for Exascale Systems

“This talk will focus on programming models and their designs for upcoming exascale systems with millions of processors and accelerators. Current status and future trends of MPI and PGAS (UPC and OpenSHMEM) programming models will be presented. We will discuss challenges in designing runtime environments for these programming models by taking into account support for multi-core, high-performance networks, GPGPUs, Intel MIC, scalable collectives (multi-core-aware, topology-aware, and power-aware), non-blocking collectives using Offload framework, one-sided RMA operations, schemes and architectures for fault-tolerance/fault-resilience.”

High Performance Computing at CSCS

“With around 3.2 billion computer operations (3.2 gigaflops) per watt, the combination of GPUs CPUs makes “Piz Daint” one of the world’s most energy-­efficient supercomputers in the petaflop performance class.”

InfiniBand Principles Every HPC Expert MUST Know

In this video from the HPC Advisory Council Swiss Conference 2014, Oded Paz from Mellanox Global Education Services presents: InfiniBand Principles Every HPC Expert MUST Know.

GTC 2014 Keynote Time Lapse from insideHPC

I produced this time lapse video featuring clips from the opening keynote at GTC 2014. “Nvidia really knows how to wow the audience with a blend of showmanship, state-of-the-art graphics, and dazzling technology.”