Optimizing Codes for Heterogeneous HPC Clusters Using OpenACC

July 3, 2017

Looking at the Top500 and Green500 ranks, one clearly realizes that most HPC systems are heterogeneous architecture using COTS (Commercial Off-The-Shelf) hardware, combining traditional multi-core CPUs with massively parallel accelerators, such as GPUs and MICs. With processor frequencies now hitting a solid wall, the only truly open avenue for riding today the Moore’s law is increasing hardware parallelism in several different ways: more computing nodes, more processors in each node, more cores within each processor, and longer vector instructions in each core. Read more…

By Enrico Calore et. al.

Compilers and More: OpenACC to OpenMP (and back again)

June 29, 2016

In the last year or so, I’ve had several academic researchers ask me whether I thought it was a good idea for them to develop a tool to automatically convert OpenACC programs to OpenMP 4 and vice versa. In each case, the motivation was that some systems had OpenMP 4 compilers (x86 plus Intel Xeon Phi Knights Corner) and others had OpenACC (x86 plus NVIDIA GPU or AMD GPU), and someone wanting to run a program across both would need two slightly different programs. In each case, the proposed research sounded like a more-or-less mechanical translation process, something more like a sophisticated awk script, and that’s doomed from the start. I will explain below in more detail how I came to this conclusion. Read more…

By Michael Wolfe

A Comparison of Heterogeneous and Manycore Programming Models

March 2, 2015

The high performance computing (HPC) community is heading toward the era of exascale machines, expected to exhibit an unprecedented level of complexity and size Read more…

By Yonghong Yan, Barbara M. Chapman and Michael Wong

New Degrees of Parallelism, Old Programming Planes

August 28, 2014

Exploiting the capabilities of HPC hardware is now more a matter of pushing into deeper levels of parallelism versus adding more cores or overclocking. What thi Read more…

By Nicole Hemsoth

Parallel Programming with OpenMP

July 31, 2014

One of the most important tools in the HPC programmer's toolbox is OpenMP, a standard for expressing shared memory parallelism that was published in 1997. The c Read more…

By Tiffany Trader

A Data Locality Cure for Irregular Applications

February 18, 2014

Data locality plays a critical role in energy-efficiency and performance in parallel programs. For data-parallel algorithms where locality is abundant, it is a Read more…

By Carlo del Mundo

Photo courtesy of Nor-Tech

Unleashing The Potential of OpenMP via Bottleneck Analysis

February 13, 2014

To capitalize on the computational potential of parallel processors, programmers must identify bottlenecks that limit their application. These bottlenecks typic Read more…

By Carlo del Mundo, Contributing Editor

The Week in HPC Research

March 21, 2013

The top research stories of the week include an evaluation of sparse matrix multiplication performance on Xeon Phi versus four other architectures; a survey of HPC energy efficiency; performance modeling of OpenMP, MPI and hybrid scientific applications using weak scaling; an exploration of anywhere, anytime cluster monitoring; and a framework for data-intensive cloud storage. Read more…

By Tiffany Trader

  • arrow
  • Click Here for More Headlines
  • arrow

Leading Solution Providers

  • Off The Wire

  • Industry Headlines

More Off The Wire

Whitepaper:

Strategies for the Spectrum of Cloud Adoptions

Whether an organization chooses a cloud for general business needs or a highly tailored workload, the spectrum of offerings and configurations can be overwhelming. To help you navigate the various cloud options available today, we're breaking down your options, exploring pros and cons, and sharing ways to keep your options open and your business agile as you execute your cloud strategy.

Download this report

Sponsored by Microsoft

Whitepaper:

Adaptive Flexibility is the Future of Supercomputing – The Arm advantage for HPC workloads

Researchers in academic labs and commercial R&D groups continue to need more compute capacity, which means leveraging the latest innovations in HPC technologies as well as an assortment of resources to meet the unique needs of different workloads. Increasingly, systems based on Arm processors are stepping into that role, offering low power consumption and strategic advantages for HPC workloads.

Download this report

Sponsored by Cray

Roundtable Webinar:

Storage Strategies in the Age of Intelligent Data

From scale-out clusters on commodity hardware, to flash-based storage with data temperature tiering, cloud-based object storage, and even tape, there are a myriad of considerations when architecting the right enterprise storage solution. In this round-table webinar, we examine case studies covering a variety of storage requirements available today. We’ll discuss when and where to use various storage media in accordance with use cases, and we’ll look at security challenges and emerging storage technology coming online.

Watch this Webinar

Sponsored by Cloudian, HPE, Pure Storage, Re-Store, and WekaIO

Advanced Scale Career Development & Workforce Enhancement Center

Featured Advanced Scale Jobs:

Receive the Monthly
Advanced Computing Job Bank Resource:

HPCwire Resource Library

HPCwire Product Showcase

Subscribe to the Monthly
Technology Product Showcase:

Subscribe