Critical issue 3: Research
How IT can support research in all disciplines,
and serve to interconnect the campus
with the greater Bay Area research community.
The original presentation of this information is available in PDF form:
Current trends in research [PDF]
A tabular presentation of this information
is available in PDF form:
Strategic technology needs
for research at UC Berkeley
[PDF]
Current trends in research
- Increased focus on interdisciplinary research by faculty and funders.
- Increased emphasis on multicampus research programs and projects.
- Increased emphasis on research collaborations with industry.
- Increased reliance on high-throughput computing.
- Increased reliance on high-bandwidth, networked computing capability.
- Increased need for sophisticated visualization and simulation and modeling software.
- Need to archive, manipulate, and mine increasingly large databases.
Sources
Presention by Vice Chancellor for Research Beth Burnside to the Berkeley
IT Architecture Committee (ITAC) about the IT needs of campus
researchers (March 1, 2005).
Bill Decker and Bonnie Neas, Research
Universities and the Central IT Organization: Rebuilding the
Partnership. EDUCAUSE Review, May/June 2003. [PDF]
Strategic technology needs
for research at UC Berkeley
Research 1:
Basic IT resources
Priority: 1 (highest)
Description
Researchers and research support
staff across disciplines require the
use of common systems or technical
compatibility, else significant
reinvestment and training is needed
for each new research initiative.
Examples of this issue
-
Departments engaging in joint
research use very different
administrative and research systems
and tools today, slowing research and
increasing costs.
-
To reduce operating costs, personal
computers and servers are often used
well beyond standard three-year life
expectancy. This results in a vast array
of aging hardware and software,
incompatible between departments.
-
Shifting from one platform to
another is very time-intensive and
costly, reducing interest in moving to
standard solutions.
Goals for this issue
-
Publish minimum standards and
baseline requirements for common
computing environments in support
of research.
-
Develop lifecycle plans to annually
evaluate baseline for coming year.
-
Provide all researchers and
research support staff (as
appropriate) with computing devices
compatible with minimum standards,
refreshed on regular schedule (bronze level).
-
Publish service catalog of available
computing environments and
standards for research computing.
Current actions taking place to meet goals
Future actions needed to meet goals
Research 2:
Technical support
Priority: 2
Description
Campus research units have
considerable hardware and software
resources which often are
administered part time, poorly
configured, and insecurely
configured.
Examples of this issue
-
Many grants provide for one-time
capital purchases but system
administration operating costs are not
covered. Student labor is often used,
irrespective of their formal training,
skills, or interest.
-
Campus models that include
recurring charges for services
(including the new data center) don't
match well with grant funding.
It is easier for the research unit
to use the grant's one-time funds to
buy its own hardware than to commit
to ongoing costs for storage.
Goals for this issue
-
Develop shared resource pool of
complementary skills that are
dedicated to supporting researchers.
-
Provide IT "boot camp" training
for all student IT employees
supporting research.
-
Create Technical Resource
Manager position to serve as a single
point of contact and identify and plan
use of technical resources across
campus.
-
Free up resources through
Resource Manager to provide
enhanced and custom service
opportunities where required.
Current actions taking place to meet goals
Future actions needed to meet goals
Research 3:
Advanced collaborative and multi-site research
Priority: 3
Description
Research endeavors between
campus units or with private sector
increasingly rely on shared resources
located at multiple institutions and
companies.
Examples of this issue
-
Today's collaboration tools at Berkeley
are extremely limited. Most
collaboration consists of email-based
file sharing. Lack of standards or
shared environments necessitates the
development of costly innovation and
duplication.
-
Private-sector collaboration often
depends on adapting to tools and
standards provided by each company
that are often incompatible with
those available to the research unit.
Goals for this issue
-
Provide open source, standard
online collaboration environments
accessible from multiple
departments, campuses, and private
enterprise.
-
Invest in next-generation tools,
including real-time collaboration,
Croquet, Sakai, and other open-source solutions.
Current actions taking place to meet goals
Future actions needed to meet goals
Research 4:
Data stewardship and digital asset management
Priority: 4
Description
Disparate, very large data sets require next-generation metadata
management solutions and application archiving practices and tools to
ensure availability of data.
Examples of this issue
-
Stewardship of research data is
essential to support researchers' needs
to archive, manipulate, and mine
increasingly large databases.
-
Collections have critical digital asset-management needs that are being
individually solved by departments at
great expense.
-
Database management is often left to
students or other nonprofessional
management, resulting in considerable
performance and security weaknesses.
-
Backups for protection of digital
assets are often done locally due to
operational cost constraints.
Goals for this issue
-
Provide options for storing very
large data sets that can be actively
accessed by multiple groups.
-
Provide backup and archive
solutions to all research projects.
-
Provide comprehensive digital
asset-management solutions,
including application archiving, to all
researchers.
-
Participate in and/or lead national
efforts to develop metadata
standards for core research areas.
Current actions taking place to meet goals
Future actions needed to meet goals
Research 5:
High-performance computing, simulation, and visualization environments
Priority: 5
Description
Computing power and research
needs double on average every 18
months, with requirements for
increased capacity of networks,
storage, and visualization tools
growing even faster.
Examples of this issue
-
Increasingly, traditional research
techniques are augmented by or replaced
with technology-based tools, methods,
and simulations. Aging computer
systems prevent timely research and
use of newest-generation tools.
-
Cross-discipline research often
requires high-speed computing. Many
of the buildings on campus do not have
sufficient physical infrastructure for
standard networking, much less
research-level bandwidth.
-
Shared research tools between
campuses or companies require
substantial bandwidth often in excess
of what is available.
-
Ultra-high-performance computing is
needed to support researchers'
demands for high-throughput and high-bandwidth
networked computing, and
sophisticated visualization, simulation,
and modeling software.
Goals for this issue
-
Provide sufficient networking and
computing resources to enable
access to remote high-performance
computing environments.
-
Regularly tested and verified
access to remotely based
supercomputing, high-performance
computing, storage, visualization,
and simulation environments.
-
Publish inventory of available campus and
systemwide research environments,
computing infrastructure, research
tools, and applications.
Current actions taking place to meet goals
Future actions needed to meet goals
