Archives for January 2017

Deep Learning in HPC – Using Data to Go Far Beyond Automation in Pathology

Dr. Amit Seti from IIT-Gauwhati presented this talk at GTCx in India. “This talk will cover how medical imaging data can be used to train computer vision systems that automate diagnostic analysis in current clinical practice. Not only that, with more creative use of data, we can go even beyond that to predict outcome of specific treatment for individual patients. We will cover results from prostate and breast cancers to show that a future is not too far where algorithms will become a necessary set of tools in a pathologist’s toolbox.”

Intel FPGAs Break Record for Deep Learning Facial Recognition

Today Intel announced record results on a new benchmark in deep learning and convolutional neural networks (CNN). ZTE’s engineers used Intel’s midrange Arria 10 FPGA for a cloud inferencing application using a CNN algorithm. “ZTE has achieved a new record – beyond a thousand images per second in facial recognition – with what is known as “theoretical high accuracy” achieved for their custom topology. Intel’s Arria 10 FPGA accelerated the raw design performance more than 10 times while maintaining the accuracy.”

Job of the Week: Wind Energy Research Engineer at NREL

“The National Renewable Energy Laboratory(NREL), located at the foothills of the Rocky Mountains in Golden, Colorado, is the nation’s primary laboratory for research and development of renewable energy and energy efficiency technologies. NREL is continuing an active research and development program for modeling of wind farm interactions and mesoscale dynamics within the National Wind Technology Center. This R&D program has an opening for one full-time engineer in wind farm modeling and mesoscale research.”

A Look at the CODAR Co-Design Center for Online Data Analysis and Reduction at Exascale

Ian Foster and other researchers in CODAR are working to overcome the gap between computation speed and the limitations in the speed and capacity of storage by developing smarter, more selective ways of reducing data without losing important information. “Exascale systems will be 50 times faster than existing systems, but it would be too expensive to build out storage that would be 50 times faster as well,” said Foster. “This means we no longer have the option to write out more data and store all of it. And if we can’t change that, then something else needs to change.”

UberCloud Obtains $1.7 Million in Pre-A Funding Round

“UberCloud has created an entire cloud computing ecosystem for complex technical simulations, leveraging cloud infrastructure providers, developing and utilizing middleware container technology, and bringing on board established and proven application software providers, all for the benefit of a growing community of engineers and scientists that need to solve critical technical problems on demand,” said Roland Manger, co-founder and Partner at Earlybird. “While technical computing has been slow to adopt the benefits of the Cloud, we are convinced that UberCloud can be a catalyst for change.”

RCE Podcast Looks at iRODS Data Management Software

In this RCE Podcast, Brock Palen and Jeff Squyres speak with the creators of iRODS: Jason Coposky and Terrell Russell. Also known as the Integrated Rule-Oriented Data System, iRODS open source data management software is used by research organizations and government agencies worldwide. “iRODS virtualizes data storage resources, so users can take control of their data, regardless of where and on what device the data is stored. The development infrastructure supports exhaustive testing on supported platforms. The plugin architecture supports microservices, storage systems, authentication, networking, databases, rule engines, and an extensible API.”

PASC17 to Feature Talk by Gordon Bell Prize Winner Haohuan Fu

“This talk reports efforts on refactoring and optimizing the climate and weather forecasting programs – CAM and WRF – on Sunway TaihuLight. To map the large code base to the millions of cores on the Sunway system, OpenACC-based refactoring was taken as the major approach, with source-to-source translator tools applied to exploit the most suitable parallelism for the CPE cluster and to fit the intermediate variable into the limited on-chip fast buffer.”

Beyond Exascale: Emerging Devices and Architectures for Computing

“Nanomagnetic devices may allow memory and logic functions to be combined in novel ways. And newer, perhaps more promising device concepts continue to emerge. At the same time, research in new architectures has also grown. Indeed, at the leading edge, researchers are beginning to focus on co-optimization of new devices and new architectures. Despite the growing research investment, the landscape of promising research opportunities outside the “FET devices and circuits box” is still largely unexplored.”

Intel Xeon Phi Processor Programming in a Nutshell

In this special guest feature, James Reinders looks at Intel Xeon Phi processors from a programmer’s perspective. “How does a programmer think of Intel Xeon Phi processors? In this brief article, I will convey how I, as a programmer, think of them. In subsequent articles, I will dive a bit more into details of various programming modes, and techniques employed for some key applications. In this article, I will endeavor to not stray into deep details – but rather offer an approachable perspective on how to think about programming for Intel Xeon Phi processors.”

IBM Adds TensorFlow Support for PowerAI Deep Learning

Today IBM announced that its PowerAI distribution for popular open source Machine Learning and Deep Learning frameworks on the POWER8 architecture now supports the TensorFlow 0.12 framework that was originally created by Google. TensorFlow support through IBM PowerAI provides enterprises with another option for fast, flexible, and production-ready tools and support for developing advanced machine learning products and systems.