ict infrastructure for science: e-science developments henri bal [email protected] vrije universiteit...
TRANSCRIPT
ICT infrastructure for Science:
e-Science developments
Henri [email protected]
Vrije Universiteit Amsterdam
Outline
● What is e-Science?● Virtual Laboratory for e-Science (VL-e)● Research infrastructure of VL-e● Some VL-e results● Future developments in the
Netherlands
Science is changing
● System level science● the integration of diverse sources of knowledge about the
constituent parts of a complex system with the goal of obtaining an understanding of the system's properties as a whole [Ian Foster]
● Multidisciplinary research● Each discipline can solve only part of a problem● Collaborations betweens distributed research
groups
● Research driven by (distributed) data● Data explosion, both volume and complexity
Examples
● Functioning of the cell for system biology● Cognition● Cancer research● Cohort studies in medicine (biobanking)● Discovery of biomarkers for drug design● Ecosystems/biodiversity● Studies of water/air pollution● Study black matter
e-Science
● Goal: allow scientists to collaborate in experiments and integration of research● Enable system level science
● Design methods to optimally exploit underlying infrastructure● Hardware (network, computing,
datastorage)● Software (web, grid middleware)
e-Science in context
Sytem levelexperiments
e-Science
Infrastructure
Web/grid software
Virtual Laboratory fore-Science (VL-e)
● 40 M€ BSIK project (2004-2009)● Generic application support
● Application cases are drivers for computer & computational science and engineering research
● Re-use of components via generic solutions
● Rationalization of experimental process● Reproducible & comparable
Optical NetworkingHigh-performance
distributed computingSecurity & Generic
AAA
Virtual lab. &System integration
Interactive PSE
Collaborative information Management
Adaptive information
disclosure
User Interfaces & Virtual reality
based visualization
Bio
-div
ers
ity
Bio
-In
form
ati
cs
Te
les
cie
nce
Dat
a In
ten
siv
eS
cie
nc
e
Fo
od
Info
rma
tics
Med
ical
dia
gn
os
is &
imag
ing
VL-e
Grid Middleware
Gigaport Network Service (lambda networking)
Application specificservice
Application Potential
Generic service &
Virtual Lab. services
Grid &
NetworkServices
Virtual Laboratory
VL-E Experimental Environment
VL-E Proof of concept Environment
Virtual Lab.rapid prototyping
(interactive simulation)
Additional Grid Services
(OGSA services)
The VL-e infrastructure
Proof-of-Concept Rapid Prototyping
(DAS-3)
DAS-3 272 nodes(AMD Opterons)
792 cores1TB memoryLAN: Myrinet 10G Gigabit
EthernetWAN: 20-40 Gb/s
OPN
● Applications can dynamically allocate light paths and change the topology of the wide-area network
● Applications: model checking, game tree search, processing CineGrid data (4K video)● Kees Verstoep’s talk (yesterday)
Grid Middleware
SurfnetNetwork Service
(lambda networking)
Virtual Laboratory
VL-E Experimental Environment
VL-E Proof of concept Environment
Rapid prototyping(interactive simulation)
Additional Grid Services
(OGSA services)
Grid Middleware
Surfnet
Virtual Laboratory
Big Grid
BiG Grid
Outline
● What is e-Science?● Virtual Laboratory for e-Science (VL-e)● Research infrastructure of VL-e● Some VL-e results
● Applications● Generic application support
(middleware)
● Future developments in the Netherlands
Functional MRI: Analysis
MR scanner
Brain activation maps
Stimulus System for Cognitive research
fMRI scan Group Activation Map
Intr
o f
MR
I
• Large datasets, many instances• Computation demanding analysis• Distributed resources (scanning, analysis)• Collaboration (data, methodology)
Medical Diagnosis and ImagingProblem Solving Environment
VL-e generic services:• Provides:
– Scientific visualization techniques– SRB– Resource browsing– Workflow management– Job submission– Data querying
• Uses:– V Browser on SRB
– Parallel processing techniques
– VLAM
Application specific services:
• Access to PACS, DICOM
• Interfaces to medical scanners (MRI)
• In-house developed algorithms:– Eddy Current Reduction
– Matched Masked Bone Elimination
• Authentication & authorization
Grid Middleware
Surfnet
Virtual Laboratory
VL-e Environment
… MedicalApplications
…
Grid services:
• Storage facilities
• High Performance Computing platforms
• High Performance Visualization
StimulusSystem
3 Tesla MRI
Dynamicbird behaviour
MODELS
Bird distributionsEnsembles
Calibration andData assimilation
Predictions and on-linewarnings
RADAR
Bird behaviour in relation to weather and landscape
Ibis – Grid programming
● Goal:● drastically simplify grid
programming/deployment
● applications running on many co-allocated resources (``grids as promised’’)
Ibis system
Ibis applications● e-Science (VL-e)
● Brain MEG-imaging● Mass spectroscopy● Grammar learning
● Multimedia content analysis● Other programming systems
● Workflow engine for astronomy (D-grid), grid file system, ProActive, Jylab, …
Multimedia content analysis
● Analyzes video streams to recognize objects
● Extract feature vectors from images● Describe properties (color, shape)● Data-parallel task implemented with C+
+/MPI
● Compute on consecutive images● Task-parallelism on a grid
MMCA
‘Most Visionary Research’ award at AAAI 2007, (Frank Seinstra et al.)
Discussion about infrastructure
● Need well-balanced infrastructure supporting compute/data/network-intensive applications
● Generic software is part of the infrastructure● Key to obtain flexibility
● Organization is important, different roles● Application experiments● Computer Science experiments● Production
● Building infrastructure is research in itself
Next: national e-Science centre?
● Coordinate e-Science research● Software services needed for e-Science● Organize support● Help in developing policies for
infrastructure