Tag Archives: cloud computing

Take a Cloud Survey From NSF + XSEDE

Posted on by

NSF and XSEDE management would like to hear directly from researchers and educators on how they are using the cloud, why they are using the cloud, and the advantages the cloud provides for their research. Researchers who are not currently XSEDE users are especially welcomed to participate.

The relatively short online survey may be found at www.xsede.org/cloudsurvey. (New users will need to register for an account.)

About the Survey
This information will help XSEDE management understand users’ cloud computing experiences so that they can better plan for integrating clouds into the XSEDE architecture. The survey results will be included in an XSEDE Cloud Study report.

Background
The goal of XSEDE is to enhance research productivity. NSF through the XSEDE integrating fabric is committed to promoting a diversity of computing resources, inclusive of clouds, and, in addition, recognizes the opportunity for cloud to play a significant role in many other parts of a scientific workflow. XSEDE must embrace cloud, identify complementary areas that cloud can support, and have a clear strategy for integrating cloud into national cyberinfrastructure. To achieve this objective, a clear understanding of use cases is needed in order to define an optimal cloud interoperability architecture for XSEDE.

Researchers need to perform computations and move data to/from the cloud and to/from other XSEDE resources with ease if XSEDE is to serve as a “one-stop” resource for a range of computation. Cloud integration issues such as authentication, data movement, and accounting, therefore, need to be defined and investigated.

Opening Science Gateways to Future Success – See What the NSF is Doing

Posted on by

In 2009, the NSF funded a study into what will make science gateways successful. Led by Katherine Lawrence of U-M, and Nancy Wilkins-Diehr of the University of California at San Diego, the study looked at:

  • the characteristics of successful gateways that warrant long-term funding, and
  • the potential for transformation in a field through new applications of gateway technologies.

Understanding the types of science and engineering problems and the types of communities that can most benefit from applied, persistent cyberinfrastructure will lead to informed investment decisions.

This 6-minute video summarizes their findings. The video was debuted at the XSEDE12 conference in Chicago, where their paper won best paper in the “Software and Software Environments” track. To read the paper, learn more about the study, and find out how you can be a part of the next steps, visit www.sciencegateways.org.

Apply for Google App Engine Award by August 31

Posted on by

The Google App Engine Education Awards program is giving grants of $1,000 in App Engine credits to individual educators. The goal of the program is to assist in and inspire knowledge exploration by offering access to the Google App Engine infrastructure to higher education faculty and students for coursework and student projects.

Award information
Awards will be given to individual educators. If selected, each professor receives $1,000 in App Engine credits which can be distributed to up to 10 applications. These credits are donated on top of the already free App Engine quotas available to all users.

Awarded Google App Engine credits can only be used for the proposed project. The award start date is flexible, but the awarded Google App Engine credits must be used within 12 consecutive months. Awardees are responsible for any usage costs that exceed the awarded amount. Unused credits will expire at the end of 12 months.

Google App Engine Education Awards are unrestricted gifts. At the completion of the project we prefer that coursework, curricula, tools and applications(s) are open sourced and made freely available to the academic community. Please see these FAQs for more program details.

Timeline
Proposals will be accepted electronically until 11:59 p.m. PST August 31, 2012. Award recipients will be notified by email.

Eligibility
Google App Engine Education Awards are open to full-time faculty or staff at an accredited academic institution in the United States, such as a college or university. Successful applicants will be required to agree to the App Engine Terms of Service.

We will fund projects that create tools, applications and curriculum that can be used by other educators in their own teaching environments. Proposed projects must support either teaching or learning activities. Below are some examples of Google App Engine uses in an educational context. These are only suggestions and are not intended to limit other creative adoptions:

Design a course that uses Google App Engine within traditional Computer Science, Math, or Science curriculum; or, a course on Programming in the Cloud that builds skills in designing for scale.
Create a teacher forum for discussing, exchanging ideas and practices, and/or improving curriculum
Build a tool for educators and school administrators to collect, track and analyze educational data
Develop a Learning Management System for course administration and management
Build a platform for learning interactively

Other examples:
University of Michigan : using Google App Engine to teach web programming courses
USF: learning to build applications on Google App Engine
University of Washington: using Google App Engine to implement social and environmental projects in emerging regions
New York University: course collaboration across the sea Services included in the Award

Google App Engine enables users to build and host web apps on the same computing infrastructure that powers Google applications.Beyond fast development and deployment, Google App Engine offers simple administration, with no need to worry about hardware, patches or backups. Besides building web apps, Google App Engine services also include automated scalability, full text search, Google Cloud SQL, Memcache, App Engine MapReduce. Your web application can run in one of three runtime environments: the Go, Java or Python environment.

More information can be found on the Google Developers website. General questions are addressed in the App Engine FAQs.

Award Extensions
After one year, an awardee may apply for an award renewal of the original project for an additional $1,000 for 12 months. Any original project may qualify. Renewal decisions will be at the sole discretion of Google and will be based on community interest, number of page views (or users), overall project progress or potential impact. All unused credits will expire at the end of the twelfth month. New projects are not eligible for an extension.

How to Apply
To apply, please fill out this form.

About App Engine
Google App Engine is an easy way to teach anyone – from first time programmers to advanced students – about complex large-scale projects. You don’t have to set up an infrastructure; your web applications run on Google’s infrastructure. There are no servers to maintain: You just upload your application, and it’s ready to serve your users. Google App Engine can be used to build mobile or social applications, traditional web browser-based applications, or standalone web services.

Important Information
Educational institutions may have ethics rules that restrict the ability of employees to participate in certain events. By submitting a proposal, the applicant certifies that the entry complies with the institution’s rules and regulations. Successful applicants must also confirm that they can accept the Google App Engine credits described in this RFP. Furthermore, non-U.S. nationals should confirm such acceptance complies with all applicable laws, including anti-bribery laws in their country of nationality. By participating in this program, you agree that all award decisions are solely within Google’s discretion. In addition, you agree not to challenge or dispute Google’s decision to award or not award Google App Engine credits. Certain tax documentation, such as an IRS Form W-9 or similar certification for non-US award recipients, will be required for anyone receiving Google App Engine credits. Please note that Google cannot accept any proposal containing confidential or proprietary information, as such all information included in the proposal will be deemed non-confidential. Google may independently conduct similar development and will own the intellectual property rights in the application(s) or content Google creates.

For questions, please contact appengineawards-education@google.com.

Science Cloud Summer School Starts July 30 – Register Now

Posted on by
Register now for the Science Cloud Summer School, which starts Monday, July 30. This session is a great set of talks and exercises, ranging from applications, to programming, to infrastructure, to workflows and education.
The Science Cloud Summer School targets education and training of graduate students and the fostering of a community around a topic that has increasing interest and relevance: the use of cloud computing technologies in science – including infrastructure-as-a-service and platform-as-a-service.
Because cloud computing systems and technologies provide a considerable departure from traditional models and evolve at a rapid pace, this event would provide a basis for students to immerse in a focused, intensive curriculum to learn fundamentals and experiment with these technologies in practice.
Date: Monday, July 30 – Friday, Aug 3
Times: 11 am to 7 pm daily
Location: HH Dow Building, 2300 Hayward Street, Ann Arbor, in classroom #3150. For more information, visit  sites.google.com/site/vscsehostsites/university-of-michigan.
Registration fee: $100
Topics:
Monday July 30 – Introduction
Tuesday July 31 – Cloud Technologies
Wednesday August 1 – Academic & Commercial Cloud Infrastructure
Thursday August 2 – Cyberinfrastructure/HPC & Clouds:  Technology & Applications
Friday August 3 – Education Applications & Advanced Technology

Virtual School for CS and Engineering Starts in July

Posted on by

Graduate and doctoral students from all disciplines can gain valuable skills in computational science — including programming for many-core processors and heterogeneous systems and leveraging cloud resources — during hands-on summer school schools offered by the Virtual School for Computational Science and Engineering (VSCSE).

The VSCSE helps students use emerging petascale computing resources to address real problems in a wide range of science and engineering disciplines.

The summer schools for 2012 are:

  • Programming Heterogeneous Parallel Computing Systems (July 10-13 2012) [GPGPU]
  • Science Cloud Summer School (July 30 – Aug 3 2012)
  • Proven Algorithmic Techniques for Many-Core Processors (Aug 13-17 2012)

These schools will be delivered to sites nationwide using high-definition videoconferencing technologies, allowing students to participate from convenient locations across the country where they will be able to:

  • work with a cohort of fellow computational scientists,
  • have access to local experts, and
  • interact in real time with school instructors.

Registration for each school is $100.

For more information, go to: www.vscse.org/summerschool/2012/ or contact:
Scott Lathrop (lathrop@illinois.edu)
Blue Waters Technical Program Manager for Education XSEDE
Director of Education and Outreach
217-714-2517

Register now at: hub.vscse.org/

Host Sites:

Programming Heterogeneous Parallel Computing Systems (July 10-13 2012)

  • U Illinois Urbana-Champaign/National Center for Supercomputing Applications
  • Harvard U, Cambridge MA
  • Louisiana State U, Center for Computation & Technology, Baton Rouge LA
  • Pittsburgh Supercomputing Center, Pittsburgh PA
  • Princeton U, Princeton Inst for Computational Science & Engineering, NJ
  • Rutgers U, Piscataway NJ
  • U California Los Angeles
  • U Michigan, Ann Arbor
  • U Oklahoma, Norman
  • U South Carolina, Columbia
  • U Tennessee Knoxville
  • U Utah, Salt Lake City UT

Science Cloud Summer School (July 30 – Aug 3 2012)

  • Indiana U, Bloomington
  • Louisiana State U, Center for Computation & Technology, Baton Rouge
  • Michigan State U, Institute for Cyber Enabled Research, East Lansing
  • Pennsylvania State U, State College
  • Princeton U, Princeton NJ
  • Rutgers U, Piscataway NJ
  • U California Los Angeles
  • U Michigan, Ann Arbor
  • U Wisconsin Milwaukee

Proven Algorithmic Techniques for Many-core Processors (Aug 13-17 2012)

  • U Illinois Urbana-Champaign, National Center for Supercomputing Applications
  • Harvard U, Cambridge MA
  • Michigan State U, Institute for Cyber Enabled Research, East Lansing
  • Pittsburgh Supercomputing Center, Pittsburgh PA
  • Pennsylvania State U, State College
  • Rutgers U, Piscataway NJ
  • U California Los Angeles
  • U Oklahoma, Norman
  • U South Carolina, Columbia
  • U Tennessee Knoxville
  • U Utah, Salt Lake City
  • Vanderbilt U, Nashville TN

 

Google Lecture Recap and Recording: Science in the Cloud

Posted on by

Dan Atkins uses his iPad to take a picture of Joe Hellerstein addressing the audience in Chesebrough Auditorium.

What does it mean to do science in the cloud, and where is it going? These questions were the topic of a public lecture delivered by Google’s manager for computational science, Joe Hellerstein. During his visit to Ann Arbor on April 26, Hellerstein addressed a crowd of 100 faculty, staff and students, providing a high-level overview of current Google cloud services and hinting at directions Google may take in the realm of computational science. Due to a blanket non-disclosure agreement with the university, Hellerstein limited his comments in the general forum, but offered to meet privately with researchers interested in hearing more details.

Prior to the lecture, Hellerstein took part in pre-scheduled meetings with select U-M researchers who are currently positioned to make immediate use of cloud tools in possible pilot collaborations.

In his talk, Hellerstein discussed how the science community can leverage Google’s experience over the last 10 years with massive data sets and building scalable infrastructure. “The starting point [of Google cloud services] may be more cycles, or cheaper cycles. But what it leads to rapidly is a fundamental change in the way we do discovery. We have an opportunity for sharing, reproducibility, and collaboration that we don’t have today with our normal mechanisms,” said Hellerstein. He went on to provide case studies of how scientists at Google are using cloud tools, and he shared his thoughts on possible opportunities that could stem from using the cloud for scientific discovery.

More than 100 faculty, staff and students gathered to learn more about Google's cloud tools for computational science.

The slides and audio from Hellerstein’s public lecture are available here.

 

 

 

 

 

 

 

CFP and Workshop on Resiliency in HPC in Clusters, Clouds, and Grids

Posted on by

The fifth Workshop on Resiliency in High Performance Computing (Resilience) in Clusters, Clouds, and Grids will be held in conjunction with the 18th International European Conference on Parallel and Distributed Computing (Euro-Par 2012) on Rhodes Island, Greece, on August 27-31, 2012.

The paper submission deadline is Friday, June 1, 2012.

Details:

Clusters, Clouds, and Grids are three different computational paradigms with the intent or potential to support High Performance Computing (HPC). Currently, they consist of hardware, management, and usage models particular to different computational regimes, e.g., high performance cluster systems designed to support tightly coupled scientific simulation codes typically utilize high-speed
interconnects and commercial cloud systems designed to support software as a service (SAS) do not. However, in order to support HPC, all must at least utilize large numbers of resources and hence effective HPC in any of these paradigms must address the issue of resiliency at large-scale.

Recent trends in HPC systems have clearly indicated that future increases in performance, in excess of those resulting from improvements in single- processor performance, will be achieved through corresponding increases in system scale, i.e., using a significantly larger component count. As the raw computational performance of these HPC systems increases from today’s tera- and peta-scale to next-generation multi peta-scale capability and beyond, their number of computational, networking, and storage components will grow from the ten-to-one-hundred thousand compute nodes of today’s systems to several hundreds of thousands of compute nodes and more in the foreseeable future.

This substantial growth in system scale, and the resulting component count, poses a challenge for HPC system and application software with respect to fault tolerance and resilience. Furthermore, recent experiences on extreme-scale HPC systems with non-recoverable soft errors, i.e., bit flips in memory, cache, registers, and logic added another major source of concern.

The probability of such errors not only grows with system size, but also with increasing architectural vulnerability caused by employing accelerators, such as FPGAs and GPUs, and by shrinking nanometer technology. Reactive fault tolerance technologies, such as checkpoint/restart, are unable to handle high failure rates due to associated overheads, while proactive resiliency technologies, such as migration, simply fail as random soft errors can’t be predicted. Moreover, soft errors may even remain undetected resulting in silent data corruption.

Important dates:

  • Paper submission deadline: June 1, 2012.
  • Notification deadline: June 29, 2012.
  • Camera ready deadline is after the workshop.

Submission guidelines:

  • Authors are invited to submit papers electronically in English in PDF format via EasyChair.
  • Submitted manuscripts should be structured as technical papers and may not exceed 10 pages, including figures, tables and references using Springer’s Lecture Notes in Computer Science (LNCS) format.
  • Submissions should include abstract, key words and the e-mail address of the corresponding author. Papers not conforming to these guidelines may be returned without review.
  • All manuscripts will be reviewed and will be judged on correctness, originality, technical strength, significance, quality of presentation, and interest and relevance to the conference attendees.
  • Submitted papers must represent original unpublished research that is not currently under review for any other conference or journal.
  • Papers not following these guidelines will be rejected without review and further action may be taken, including (but not limited to) notifications sent to the heads of the institutions of the authors and sponsors of the conference.
  • Submissions received after the due date, exceeding length limit, or not appropriately structured may also not be considered. The proceedings will be published in Springer’s LNCS as post-conference proceedings.
  • At least one author of an accepted paper must register for and attend the workshop for inclusion in the proceedings.
  • Authors may contact the workshop program chair for more information.

Topics of interest include, but are not limited to:

  • Reports on current HPC system and application resiliency
  • HPC resiliency metrics and standards
  • HPC system and application resiliency analysis
  • HPC system and application-level fault handling and anticipation
  • HPC system and application health monitoring
  • Resiliency for HPC file and storage systems
  • System-level checkpoint/restart for HPC
  • System-level migration for HPC
  • Algorithm-based resiliency fundamentals for HPC (not Hadoop)
  • Fault tolerant MPI concepts and solutions
  • Soft error detection and recovery in HPC systems
  • HPC system and application log analysis
  • Statistical methods to identify failure root causes
  • Fault injection studies in HPC environments
  • High availability solutions for HPC systems
  • Reliability and availability analysis
  • Hardware for fault detection and recovery
  • Resource management for system resiliency and availability