KAREN wiki

KIWI ADVANCED RESEARCH AND EDUCATION NETWORK

E-Research

From KAREN wiki

Jump to: navigation, search

Contents

Recent news: BeSTGRID and BlueFern develop a collaborative vision for High Performance Computing and eResearch for New Zealand

New Zealand’s research community requires increasingly sophisticated software tools, computational platforms and massive storage to be successful. But creating and sustaining such an infrastructure is beyond the financial means of any single organisation. Like other advanced science nations, we need a coordinated, national approach rather than piecemeal, local funding. BeSTGRID and BlueFern have agreed to work together to develop plans to build this infrastructure, and are now in the process of applying for the funds needed for the creation and operation phases.

The RIAG Proposal below, led by Profs. Mark Gahegan (BeSTGRID and Auckland and Tim David (BlueFern and Canterbury), is currently under consideration by MoRST’s Research Infrastructure Advisory Group (RIAG). It describes a coherent ecosystem of High Performance Computing and eResearch tools and services, to bring advanced computing to the desktop of our science community.

We would value feedback on the ideas and are currently seeking expressions of support from researchers and institutions around the country. Please do get in touch if you have ideas, comments or want to know how you can support this venture.

RIAG HOC eResearch Proposal (2009)

The need for e-Research & HPC Outreach and Education in New Zealand

Mark Gahegan, Nick Jones & Tim David “A healthy digital economy is one in which information systems are constantly adapting to the changing needs of government, business and citizens. The health of a digital economy relies upon a strong supply of architects, managers and business leaders capable of understanding not only the information systems, but also the context in which they are embedded.” Paul Jeffreys, U. Oxford, 2008.

University of Queensland eResearch Video

Synopsis of the current situation

As e-Research and High Speed Computing (HPC) in New Zealand struggles to expand out from its established high speed networking and power requirements to meet the science needs of its top researchers, there are many challenges still to overcome, and many benefits yet to be realised. With a first round of funding to establish a national e-Research network (the Kiwi Advanced Research and Education Network, KAREN) and the TEC Funded BeSTGRID Project (2006-2008), we have some national capacity for collaborative science. In addition NZ now has, through Blue Fern (www.bluefern.canterbury.ac.nz), access to state-of-the-art HPC technology.

However, these benefits are in danger of being lost because: (i) the research capabilities operating on these infrastructures are as yet in too fragile a state to be fully self-sustaining, relying on just a few key individuals/institutions and with little awareness of the broader issues and opportunities within the wider research community, (ii) the infrastructure developed is not yet sophisticated enough to support the wealth and depth needed to place the New Zealand research community on an equal footing with the rest of the developed world, and (iii) the funding for collaborative research infrastructure is unstructured and non strategic in focus (short term, project oriented)

A longer term vision, and associated action, is needed to truly reap the rewards of these investments in shared infrastructure. Such a vision would need to be compelling and clearly and strongly articulated to secure an ongoing and higher level of strategic investment. Without a national strategy and long term prognosis, it will be difficult to convince researchers, educators and their administrators to take seriously the capabilities developed so far. Appendix 1 provides a very brief overview of the science, societal and economic arguments for why eResearch & HPC needs to succeed in New Zealand, along with noting what we currently have, what is needed, and some envisioning statements to set up aspirations. We suggest that building a literate and engaged community is the first step towards articulating this vision fully at a nationally level. This vision needs to show direct benefits to researchers by showcasing success stories, opportunities and capabilities to hand.

We propose here to offer two programs of work that will lead to greater awareness and understanding of eResearch & HPC nationally, will foster new collaborations, and will form the basis for establishing a coordinated approach to development of a national eResearch / HPC strategy.

The first is a National Outreach and Coordination Programme for eResearch & HPC, targeted primarily at the higher-education and research institutions (universities, CRIs). The second is a Nationally Coordinated eResearch & HPC Education Programme, run over KAREN, using appropriate technologies. National eResearch/HPC Outreach and Coordination Programme

National eResearch/HPC Outreach and Coordination Programme

Few of the universities and CRIs are able to dedicate resources to the construction of cyberinfrastructure, or to performing outreach to their own constituents with potential to apply eResearch and utilise HPC to its fullest extent.

We propose to build community nationally through a series of outreach events, targeting senior management, early adopters and potential users of eResearch and HPC from among the universities and CRIs.

The expected outcomes here would be as follows:

  1. A new road-map of national eResearch & HPC take-up and potential.
  2. A scan of emerging eResearch & HPC technologies and their potential to enable collaboration in research and higher education.
  3. Greater awareness among senior management and scientists of the needs to engage in eResearch & HPC, of the benefits already available via shared infrastructure such as KAREN and BeSTGRID, and the need to sustain these efforts, along with awareness of the organisational and other impediments encountered, that currently hinder eResearch & HPC.
  4. The foundations for a nationally-coordinated strategy for eResearch & HPC, involving all major constituents.
  5. Specific eResearch & HPC workshop(s) for key science and developer communities (in collaboration with the ARCS group in Australia). These will be targeted at increasing uptake of eResearch and HPC by providing experience and training with specific tools and technologies.

Note that the proposers have already begun this task locally in a limited capacity, working with IRL, the MaCDiarmid Institute, HORT Research, Scion and Landcare’s SCENZGrid group, and coordinating related workshops and HPC satellite meetings at the APAN meeting in Queenstown and the eResearch Australasia 2008 meeting in Melbourne.

If this work is funded, it is our intention to broaden our efforts to include other KAREN, eResearch and HPC champions from additional organisations, building a national team devoted to the promotion of eResearch / HPC and other collaborative interactions that are enabled by high-speed networks. This group would help guide REANNZ and the research community in the evolving use of network-based technologies in enabling research, and would help establish the research use-cases that might help to justify ongoing support for REANNZ.

National Outreach

There is also a need to share our eResearch / HPC experience and knowledge nationally. To this end, we will organise a national education and awareness campaign, including a National eResearch / HPC Seminar Series, to be delivered via KAREN, during the academic year 2009, and eResearch / HPC workshops organised to coincide with appropriate local conferences, and in collaboration with the Australian ARCS programme. The Australians are keen to participate with us in this venture. Our contact with ARCS is Jim McGovern (Jim.McGovern@arcs.arcs.org.au).

The series will promote the following outcomes:

  1. Establishment of community in a variety of disciplines that are interested in progressing collaborative and shared eResearch & HPC infrastructure nationally.
  2. Educate researchers, graduate students and tertiary educators by drawing on best practice and leading edge research from both offshore and NZ-based eResearch & HPC leaders.
  3. Facilitate sharing of experience, ideas and developed tools and services amongst the local NZ community.
  4. The beginnings of an eResearch & HPC educational programme and collections of resources for teaching & learning that can form the basis of a bid to TEC for more sustained, ongoing support of education in this area.
  5. Help to develop a general awareness of the benefits and possibilities offered by eResearch & HPC.

It is hoped that, over time, seminar series such as this might form the basis of an ongoing educational programme in eResearch (such as is being pursued at top universities in Europe and the USA). In that regard this seminar series could be characterised as a trial or prototype to assess the need for and feed into the design and development of a longer running seminar series, and perhaps even a number of post-graduate diplomas or Masters. Options to develop a recognised national education programme in eResearch / HPC, with support from TEC, are currently under discussion.

Broader Outcomes

We anticipate the following broader outcomes in addition.

  1. Better coordination and uptake of eResearch / HPC activities nationally.
  2. Increased awareness of eResearch / HPC applications, services, and technologies,
  3. Increased awareness of strategic models for coordinating eResearch / HPC at a national level.
  4. Maintenance of wiki content and a report detailing the findings of the outreach activities, the key players, the impediments, and the potentials (hosted from within the KAREN infrastructure), with focus on selected domains of science.
  5. A recorded series of high quality eSeminars that can be used as a community resource for educational purposes.

What we need

  • Funds to cover travel to visit with groups of researchers, senior managment and administrators at tertiary institutions and CRIs (airfares, accommodation, meals, taxis). Follow-up meetings and interaction may also be held over KAREN.
  • Funding for those involved in site visits and road shows.
  • Funding to devote to the planning and logistics for the eResearch / HPC Seminar Series, presentations and conference workshops.

Timeline

Process / Structure
  • Preparation early 2009 for a second quarter start
  • Periodic events in varying locations, commensurate with funding level
  • Communication on event activities and outcomes out to wider group
  • Ongoing conversations during the course of the year to build community and evolve related discussion among key eResearch practitioners, university and CRI managers, senior researchers
  • Creation of compelling use-cases that highlight direct science advantage

Funding Request

We request that the CBF panel consider this plan to invest some part of their remaining funds.

If there is a favourable response we would be delighted to provide a detailed budget and workplan for the above tasks. Tasks and priorities will be set according to the funds made available.


Appendix 1. The need for e-Research in NZ (science, society, economy)

Science argument for e-research

  • Help address previously intractable questions (archaeology, epidemiology, plate tectonics, astrophysics, bioinformatics, bioengineering, complex fluids engineering)
  • There is a need/obligation for NZ to share its science internationally
  • Keep a better record of our data, our experiments, our knowledge, for future scientists.
  • Current science records are woefully inadequate for deep, long-term understanding.
  • Enable more effective collaboration between experts, within NZ & internationally

Societal argument for e-Research

  • Interoperable products and services between research agencies allow for anticipation, better planning, mitigation of pressing issues such as extreme events, climate changes, invasive species, bio-security
  • Better access by educators and general public to science products & government information (e-government)
  • NZ researchers get good training in emerging science practices & do not become isolated

Economic argument for e-Research

  • Help keep NZ’s top researchers in NZ (being there is everything)
  • Promote efficiencies in the process of science and in research funding by providing shared access to expensive equipment
  • Create a culture of sharing outcomes, data, tools early and often within the science community by providing a more coordinated infrastructure

What do we have, what do we need, and how to proceed What do we have?

  • A good high-speed network with capacity for growth
  • The foundations of a sophisticated infrastructure, some HPC and massive storage silos (within a few institutions),
  • An embryonic Identity and Access Management (IAM) framework and community
  • Some grid middleware and basic services (within a few institutions)
  • Good links to e-research and HPC development communities: e.g. Australian National Collaborative Research Infrastructure Strategy (NCRIS), San Diego Supercomputer Center, IBM Research Centres (Almaden and TJ Watson), Argonne National Laboratories, eScience Edinburgh, PRAGMA, APAN, Globus.

What do we need?

  • Stay in touch with the global and national scale research & development efforts underway offshore
  • Play a role in international cyber-infrastructures, HPC & e-Research projects
  • Migrate established researchers to these new technologies
  • Train the next generation of NZ scientists
  • A detailed plan to make e-Research and HPC sustainable over the next 5-10 years
  • Plans for curation and uptake strategies for massive digital collections over the
Retrieved from "http://www.wiki.karen.net.nz/index.php/E-Research"