Copy
 What's New?
Research Highlights
Events




Research Computing + Cyberinfrastructure (RCCI) Maintenance Updates 

RCCI completed many maintenance activities this month with the exception of migrating the HPC home directories and a few other volumes to the new Panasas AS20 storage system. 

Highlights of Completed Activities:

  • Installation and validation of the new Panasas AS20 storage system
  • Upgrade of Research Data Storage NAS appliances
  • Operating system patching and updates to the cluster systems
  • Migration of compute nodes from Redcat to Rider


RCCI will be scheduling another downtime this Spring to complete the migration of the HPC home directories and a few other volumes. RCCI thanks users for their patience during this transition.

REDCap Updates

REDCap was successfully migrated into the Secure Research Environment (SRE) earlier this month. At this time, there is no change in access or service. Questions? Please contact the [U]Tech Service Desk at help@case.edu or (216)368-HELP

The next phase of REDCap includes the [U]Tech Information Security Office staff contacting PIs who may have identifiable data to evaluate options for additional security controls. This would include projects that may contain Protected Health Information (PHI) from HIPAA covered entities, Personally Identified Information (PII), or new users or projects migrating from partnering hospital systems (with PHI or PII present). In these cases, enhanced security options will be available and UTech/REDCap staff will be available for assistance.

HPC Updates

A friendly reminder that REDCat will be decommissioned in June 2018. Users need to make sure that their programs and research have transitioned to Rider. For assistance, please contact hpc-support@case.edu.

Back to Top
Want to Contribute?
If you have suggestions for items that you would like to see included in the newsletter, please send them to rcci-newsletter@case.edu or fill out our
Google form.

Effect of Meltdown and Spectre on the Performance of HPC Applications

Earlier this month, members of the CWRU community were made aware of major flaws in computer chips that could leave a huge number of computers and smartphones vulnerable to security concerns. These flaws know as SPECTRE and MELTDOWN represent issues with a system's CPU allowing an unprivileged attacker to bypass conventional memory security restrictions to gain read access to privileged memory that would otherwise be inaccessible. Both flaws require that an attacker have local system access to attempt to execute these exploits. Typically, HPC systems, including RedCat and Rider, allow multi-user access, which made these flaws of particular interest to RCCI and others in the HPC community. 

RCCI's colleagues at the Center for Computational Research at the University of Buffalo published a manuscript titled “Effect of Meltdown and Spectre Patches on the Performance of HPC Applications” that explores the performance of real world applications including NAMD and NWChem.  Based on their analysis utilizing a small development HPC cluster that was relatively isolated, the team observed that while “some specific functions can have performance decreased by as much as 74%, the majority of individual metrics indicates little to no decrease in performance. The real-world applications show a 2-3% decrease in performance for single node jobs and a 5-11% decrease for parallel multi node jobs.”

Based on this information, RCCI is in the process of reviewing the impact on our production HPC systems. In order to partially mitigate the risk of MELTDOWN and SPECTRE, RCCI applied patches to both login and management node types during the last maintenance window as recommended by various vendors. RCCI is currently evaluating the impact of performance against the security posture of our compute node systems. At this time, there is no known malicious program available to take advantage of these flaws.  

RCCI expects that the IT industry as a whole will be addressing these concerns for quite some time. RCCI and UTech will continue to closely monitor future announcements of our various vendors and information security partners indicative of potential exploitation until all patches have been applied and benchmark performance before and after patching activities are completed. Questions? Please contact us at hpc-support@case.edu.

Back to Top

Events


Working with the HPC - Installing Software - February 6

Register Here

Date: Tuesday, February 6

Location: Toepfer Room, Adelbert Hall

Time: 10:00 a.m. - Noon

Course Materials

In this workshop, researchers will learn how to install software on Rider and how to create user modules. Participants will see how to install R modules, Python modules with "setup" and "pip", and how to compile software using configure—make—make install.

  • R modules
  • Python modules
  • Compilation of binaries and libraries
  • User modules

Working with the HPC - Data Transfer - February 20

Register Here

Date: Tuesday, February 20

Location: Toepfer Room, Adelbert Hall

Time: 2:00 p.m. - 4:00 p.m.

Course Materials

This event will present strategies and techniques to work more productively on the HPC cluster. The topics will include bash scripts, cron setup, batch permission change and others concerned with efficiency and automation of tasks.


Working Amazon Web Services - AWS for Researchers - March 6

Register Here

Date: Tuesday, March 6

Location: Toepfer Room, Adelbert Hall

Time: 10:00 a.m. - Noon

Course Materials

This session will present features and capabilities of Amazon Web Services that might be of interest to researchers.


Working with the HPC - Linux Command Line - Mar. 21

Register Here

Date: Wednesday, March 21

Location: Toepfer Room, Adelbert Hall

Time: 2:00 p.m. - 4:00 p.m.

Course Materials

This event will present strategies and techniques to work more productively on the HPC cluster. The topics will include bash scripts, cron setup, batch permission change and others concerned with efficiency and automation of tasks.


Working with Caffe, Theano and Torch - Applications of Deep Learning - April 3

Registration Here

Date: Tuesday, April 3

Location: Toepfer Room, Adelbert Hall

Time: 2:00 p.m. - 3:00 p.m.

Course Materials

This lab introduces the rapidly developing technology of deep learning accelerated by GPUs. The course is intended for anyone looking for a fundamental understanding of deep learning.

In this lab, attendees will learn:

  • The concept of deep learning
  • How the growth of deep learning has improved machine perception tasks including visual perception, speech recognition and natural language
  • How to choose which software framework best suits your needs

Working with Image Classification - NVIDIA DIGITS - April 3

Registration Here

Date: Tuesday, April 3

Location: Toepfer Room, Adelbert Hall

Time: 3:00 p.m. - 4:00 p.m.

Course Materials

This lab shows participants how to leverage deep neural networks (DNN) - specifically convolutional neural networks (CNN) - within the deep learning workflow to solve a real-world image classification problem using NVIDIA DIGITS on top of the Caffe framework and the MNIST hand-written digits dataset.

In this lab, attendees will learn how to:

  • Architect a Deep Neural Network to run on a GPU
  • Manage the process of data preparation, model definition, model training and troubleshooting
  • Use validation data to test and try different strategies for improving model performance

On completion of this lab, participants will be able to use NVIDIA DIGITS to architect, train, evaluate and enhance the accuracy of CNNs on their own image classification application.

Back to Top
Copyright © 2017, All rights reserved.
Case Western Reserve University

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list