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


Video: NVIDIA Accelerates Conversational AI

In this video, NVIDIA’s Bryan Catanzaro explains how recent breakthroughs in natural language understanding bring us one step closer to conversational AI. “Today NVIDIA announced breakthroughs in language understanding that allow businesses to engage more naturally with customers using real-time conversational AI. “NVIDIA’s groundbreaking work accelerating these models allows organizations to create new, state-of-the-art services that can assist and delight their customers in ways never before imagined.”

Video: Providing a National Software Repository for Interactive HPC

Lance Wilson from Monash University gave this talk at the Singularity User Group. “The repository for the build files is public and is run in the same way as a software development project. In addition to the build files being public, the repository is linked to singularity hub, such that the containers are easily available to anyone in the characterisation research community.”

OSS Steps Up with AMD EPYC 7002 Series Processor for ‘AI on the Fly’

One Stop Systems just unveiled its comprehensive portfolio of PCIe Gen 4 expansion building blocks supporting servers based on the new AMD EPYC 7002 processor. “We are pleased to work with AMD to lead the market transition to PCIe Gen 4 by delivering flexible and scalable expansion platforms complementary to AMD EPYC 7002 based servers.”

Video: High-Performance Computing with Python – Reducing Bottlenecks

This course addresses scientists with a working knowledge of NumPy who wish to explore the productivity gains made possible by Python for HPC. “We will show how Python can be used on parallel architectures and how to optimize critical parts of the kernel using various tools. The following topics will be covered: – Interactive parallel programming with IPython – Profiling and optimization – High-performance NumPy – Just-in-time compilation with Numba – Distributed-memory parallel programming with Python and MPI – Bindings to other programming languages and HPC libraries – Interfaces to GPUs.”

OSS Awarded $36 Million Contract to Provide Flash Storage Arrays for Military Radar Application

Today One Stop Systems announced a five-year sole source agreement valued at $36 million to provide flash storage arrays to a prime contractor for the U.S. Navy. The systems will be used within state-of-the-art Navy surveillance aircraft deployed worldwide. “Under the agreement, OSS will supply full mil-spec flash storage systems that include removable canisters, allowing collected data to be removed from the aircraft and transferred to ground stations. The systems store real-time data collected from advanced airborne sensors (AAS), including radar and other sensors. The contract includes airborne and ground systems, spare canisters and support services.”

Lenovo Launches Single Socket Servers with AMD EPYC 7002 Series

Today Lenovo introduced the Lenovo ThinkSystem SR635 and SR655 server platforms, two of the industry’s most powerful single-socket servers. “Based on next-generation AMD EPYC 7002 Series processors, Lenovo’s new ThinkSystem SR635 and SR655 server platforms not only allow customers to run more workloads on fewer servers, but also offer up to 73 percent savings on potential software licensing, empowering users to accelerate emerging workloads more efficiently.”

GigaIO Extends Next-Generation Network to Storage Systems

Today GigaIO introduced the FabreX implementation of Non-Volatile Memory Express over Fabrics (NVMe-oF) architecture, streamlining NVMe network communication and large-scale storage sharing with industry-leading low latency and high-bandwidth features. “With FabreX and NVMe-oF, storage located in a server, directly attached to a server and attached across a network delivers identical performance, allowing customers to choose their preferred storage architectures and experience the same quality of service to meet desired service level objectives.”

Professor Michela Taufer to Receive 2019 IBM Faculty Award

Professor Michela Taufer from the University of Tennessee has been selected to receive a 2019 IBM Faculty Award for $20,000 USD. This award is highly competitive and recognizes her leadership in High Performance Computing and its importance to the computing industry. “Dr. Taufer will lead work with IBM to bring the IBM Onsite Deep Learning Workshop on campus. As Deep learning and AI gain traction in scientific discovery, training EECS students in these disciplines will open tremendous opportunities for them in Tennessee and outside the state.”

Interactive Supercomputing with Jupyter and DASK

Anderson Banihirwe from NCAR gave this talk at SciPy 2019. “This talk demonstrates how to use Dask and Jupyter on large high-performance computing (HPC) systems to scale and accelerate large interactive data analysis tasks — effectively turning HPC systems into interactive big-data platforms. We will introduce dask-jobqueue which allows users to seamlessly deploy and scale dask on HPC clusters that use a variety of job queuing systems such as PBS, Slurm, SGE, and LSF. We will also introduce dask-mpi, a Python package that makes deploying Dask easy from within a distributed MPI environment.”

Designing Future Flash Storage Systems for HPC and Beyond

Glenn Lockwood from NERSC gave this talk at the Samsung Forum. “In this talk, we will compare the storage and I/O requirements of large-scale HPC workloads with those of the cloud and show how HPC’s unique requirements have led NERSC to deploy NVMe in the form of burst buffers and all-flash parallel file systems rather than block- and object-based storage. We will then explore how recent technological advances that target enterprise and cloud I/O workloads may also benefit HPC, and we will highlight a few remaining challenge areas in which innovation is required.”