Irena Bojanova - Home
Cloud Computing in Research and Education
Irena Bojanova
NOV 05, 2013 10:45 AM
A+ A A-
painting of cowboy wrangling wild horses

It's the High-Tech Wild, Wild West out there!

 Although the Cloud Computing marketplace is still chaotic, it is:

  • Exciting
  • Fast-growing
  • Full of opportunities

Cloud computing has huge potential to accelerate research, enhance collaboration, and enrich education. Educators, research administrators, IT directors, and research should realize and leverage cloud’s potential in research, teaching, and learning. They can complement their current in-house cyberinfrastructure by deploying public, private, or hybrid clouds. In fact, as a recent survey reveals, some are already benefiting by embracing the clouds.

The Extreme Science and Engineering Discovery Environment (XSEDEsurvey, sponsored by the National Science Foundation (NSF),  reveals how cloud is used across a wide variety of scientific fields and the humanities, arts, and social sciences, and illustrates the potential of cloud in accelerating research, enhancing collaboration, and enriching education.  The survey conducted from September 2012 to April 2013 received responses from 80 cloud users around the globe. The XSEDE Cloud Survey Report  showcases how cloud computing services (computational and sharing capabilities) currently are - and can be – used for doing research in various fields. It also summarizes extensive data on core usage, preferred storage, bandwidth, etc., and discusses cloud benefits and limitations for specific use cases.

A special feature of the survey is, unlike most cloud surveys conducted to date, it is focused solely on the use of clouds for research and education rather than administrative or business IT. The survey report presents several helpful key finding, including top reasons to use cloud, applications that are good for cloud, and benefits and challenges of cloud. For a concise summary of the findings, please refer to Table 1.

Table 1. Key Findings on Cloud Computing in Research and Education.                                                         

Key Findings


Top reasons to use Cloud

  • On-demand access to burst resources
  • Compute and data analysis support for high throughput scientific workflows
  • Enhanced collaboration through rapid deployment of research web sites and data sharing.

Applications Good for Cloud

  • MapReduce – for processing and analyzing large data sets.
  • High throughput, embarrassingly parallel workloads – for analyzing thousands of molecules,
  • particle collisions, etc.
  • Academic labs and teaching tools – for scaling educational experiences to up to thousands of students. Cloud-based labs are either always on or provisioned on-demand.
  • Domain-specific computing environments – Science as a Service provides rich web applications and platform components that reduce time to science.
  • Commonly requested software – Software as a Service (SaaS) environments provide researchers and educators with economies of scale in software licenses and more optimal execution environments.
  • Science Gateways – rapid elasticity of cloud-based gateways supports large communities with on-demand services.
  • Event-driven science – applications that must scale quickly to respond to real-time events.

Cloud Benefits

  • Pay as you go, compute elasticity, data elasticity.
  • Can reduce capital expenditures, and associated operation and maintenance costs.
  • SaaS, Education as a Service – small labs, departments, and budget-constrained schools can access otherwise unavailable computing capabilities.
  • Broader use – increased number and diversity (including underrepresented groups) of researchers, educators, and students participating as creators and users of cyberinfrastructure.
  • Scientific workflows, rapid prototyping, data analysis

Cloud Challenges

  • Learning curve – creating, deploying, and managing a cloud instance.
  • Virtual machine performance – performance of applications may be not optimal.
  • Bandwidth, memory limits, database instability, private/public cloud interoperability, storage
  • Data movement costs – service providers charge by GB (so, researchers generate data in the cloud and leave it there or use community data sets).
  • Cloud computing cost and the funding availability.
  • However, researchers and educators have no big concern about privacy

Continued Investment Needed

  • Basic, applied, and experimental cloud computing research
  • Access to production cloud resources, cloud training, and cloud user consulting
  • CC research: domain-specific applications, dynamic provisioning of images, network support for clouds, data portability, and aggregating heterogeneous resources as services
  • CS research: cloud-hosted real-time intelligence systems, multiparty security dataflow solutions for OpenFlow networks, and big-data machine learning algorithms for rapidly evolving data sets
  • Interest in Multi-clouds -- most private clouds are expected to become hybrid clouds; challenge will be implementing a management framework that can span all cloud environments.

Source: XSEDE Cloud Survey Report.

It is notable that the survey participants found MapReduce was the most heavily used special feature offered by the cloud service providers, followed by access to community datasets.

The survey identified 12 cloud use cases categories: burst resources; collaboration; commonly requested software; computer science research; computing and data analysis support for scientific workflows;  data archiving;  data management and analysis; data sharing; domain-specific computing environments;  education, outreach, and training (EOT); event-driven real-time science; and science gateways. The survey participants were asked to select which cloud use cases their research or education project represented. Burst resources was cited as the most common cloud use case (43%), followed by computing and data analysis support for scientific workflows (35%), collaboration (35%), data sharing, and data management and analysis. Education, outreach, and training (EOT) and the use of the cloud for computer science research were also commonly cited use cases.

The researchers and educators surveyed used a variety of public and private cloud service providers.  They identified 10 key cloud benefits and 10 key challenges in using cloud.

The survey also reveals that:           

  • A more comprehensive and balanced cyberinfrastructure, i.e., a multi-level CI, is needed to support the entire spectrum of NSF-funded communities.
  • Unlike traditional HPC workloads, many of the research and education applications surveyed required many cores rather than fastest performance per core.
  • The challenges of using the cloud require continued investments in basic, applied, and experimental research.
  • Investments that facilitate access to production cloud resources, cloud training, and cloud user consulting are needed as well, whether clouds are public, private, or national cyberinfrastructure, or, more likely, some combination thereof.
  • Although in their infancy, hybrid clouds hold the promise of enabling modest size private clouds used for steady-state workloads to burst to public, community, or national cyberinfrastructure during peak workloads. Most private clouds are expected to become hybrid clouds in the future. The challenge will be implementing a management framework.

This report will help “educators, research administrators, CIOs, and research computing practitioners envision what role cloud might play in research, teaching, and learning at their respective institutions.”

Please share your thoughts on, and experiences in, using cloud computing for research and education.

Irena BojanovaIrena Bojanova, Ph.D., is a General Chair of IT Professional Conference, Editor of Encyclopedia of Cloud Computing, Wiley (to be published 2014), and the Founding Chair of IEEE CS Cloud Computing STC. She is also an Associate Editor in Chief and the Editor of the Trends Department of IEEE IT Professional, an Associate Editor of IEEE Transactions on Cloud Computing, and an Associate Editor of International Journal of Big Data Intelligence (IJBDI). Dr. Bojanova is a professor and program director, Information and Technology Systems, at University of Maryland University College, managed academic programs at Johns Hopkins University and PIsoft Ltd., and co-started OBS Ltd., (now CSC Bulgaria). She is a senior member of IEEE and can be reached at


[%= name %]
[%= createDate %]
[%= comment %]
Share this:
Please login to enter a comment: