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


New Paper Surveys Optimization Techniques for Intel Xeon Phi

A new paper by Dr Sparsh Mittal surveys techniques for evaluating and optimizing Intel’s Xeon Phi. Now accepted in Concurrency and Computation 2020, the survey reviews nearly 100 papers. “Intel Xeon Phi combines the parallel processing power of a many-core accelerator with the programming ease of CPUs. Phi has powered many supercomputers, e.g., in June 2018 list of Top500 supercomputers, 19 supercomputers used Phi as the main processing unit. This paper surveys works that study the architecture of Phi and use it as an accelerator for various applications. It critically examines the performance bottlenecks and optimization strategies for Phi. For example, the main motivation and justification for development of Phi was ease of programming.”

SLIDE algorithm for training deep neural nets faster on CPUs than GPUs

Rice researchers created a cost-saving alternative to GPU, an algorithm called “sub-linear deep learning engine” (SLIDE) that uses general purpose central processing units (CPUs) without specialized acceleration hardware. “Our tests show that SLIDE is the first smart algorithmic implementation of deep learning on CPU that can outperform GPU hardware acceleration on industry-scale recommendation datasets with large fully connected architectures.”

Argonne Publishes AI for Science Report

Argonne National Lab has published a comprehensive AI for Science Report based on a series of Town Hall meetings held in 2019. Hosted by Argonne, Oak Ridge, and Berkeley National Laboratories, the four town hall meetings were attended by more than 1,000 U.S. scientists and engineers. The goal of the town hall series was to examine scientific opportunities in the areas of artificial intelligence (AI), Big Data, and high-performance computing (HPC) in the next decade, and to capture the big ideas, grand challenges, and next steps to realizing these opportunities.

Panasas ActiveStor Solution: Architectural Overview

Panasas has released this timely new white paper “Panasas ActiveStor Solution: Architectural Overview.” The Panasas ActiveStor architecture running the PanFS storage operating system breaks through the performance constraints of other parallel file systems. The comprehensive and tightly integrated solution enables high-performance direct parallel access to petabytes of data while avoiding the stability problems inherent in legacy NAS systems as they grow.

UberCloud Publishes Compendium Of Case Studies in Life Sciences

If you are considering moving some of your HPC workload to the Cloud, nothing leads the way like a good set of case studies in your scientific domain. To this end, our good friends at the UberCloud have published a compendium entitled, Exploring Life Sciences in the Cloud. The document includes 36 CFD case studies summarizing HPC Cloud projects that the UberCloud has performed together with the engineering community over the last six years. “From the 220 cloud experiments we have done so far, we selected 15 case studies related to the life sciences. We document the results of these research teams, their applications, findings, challenges, lessons learned, and recommendations.”

The GigaIO FabreX Network – New Frontiers in Networking For Big Data

GigaIO has developed a new whitepaper to describe GigaIO FabreX, a fundamentally new network architecture that integrates computing, storage, and other communication I/O into a single-system cluster network, using industry standard PCIe (peripheral component interconnect express) technology.

Whitepaper: Accelerate Training of Deep Neural Networks with MemComputing

“The paper addresses the inherent limitations associated with today’s most popular gradient-based methods, such as Adaptive Moment Estimation (ADAM) and Stochastic Gradient Descent (SGD), which incorporate backpropagation. MemComputing’s approach instead aims towards a more global and parallelized optimization algorithm, achievable through its entirely new computing architecture.”

Compendium of articles published on Numerical Algorithms for HPC Science

The Royal Society Publishing has recently released a special compendium of articles based on a recent scientific discussion meeting with HPC Industry thought leaders. “This issue contains contributions from those who develop and implement numerical algorithms and software libraries – numerical analysts, computer scientists, and high-performance computing researchers – with those who use them in some of today’s most challenging applications.”

vScaler Launches AI Reference Architecture

A new AI reference architecture from vScaler describes how to simplify the configuration and management of software and storage in a cost-effective and easy to use environment. “vScaler – an optimized cloud platform built with AI and Deep Learning workloads in mind – provides you with a production ready environment with integrated Deep Learning application stacks, RDMA accelerated fabric and optimized NVMe storage, eliminating the administrative burden of setting up these complex AI environments manually.”

ECP Report: Advancing Scientific Productivity through Better Scientific Software

The Exascale Computing Project has published a new report to foster and advance software productivity and sustainability for extreme-scale computational science. The report introduces work by the IDEAS-ECP project, explaining its approach, outcomes, and impact of work in partnership with the ECP and broader computational science community. The DOE Exascale Computing Project (ECP) provides a […]