N1 Grid Solution

1
N1 Grid Solution

• Policy Management
• Owners negotiate usage
• 4 different, customizable policy schemes
• Exceptions for specific needs
• What are the benefits
• Equitable, enforceable sharing between groups
• Alignment of resources with business goals

2
Industry Application Interdependence
DRMAA
3
Login Screen N1 Grid Engine Portal
4
Grid Computing Components
Large SMPs
Small SMPsThin Servers and Blades
Storage
Workstationsand Thin Clients
More than just clusters More than just connected
storage, compute, and display systems
5
Data Grid SAN, NAS, DASStorage Access,
Management and Archival
UNIX Servers (many)
Windows Linux Servers
3rd PartySwitches
StorEdge 3000Low Cost per GB
StorEdge 9000Low ComplexityHigh Performance
StorEdge 6000Modular Flexibility
6
Graphics Grid Access for More Users to
Visualization Services atRequired Visual Quality
and Performance Levels
Visualization
Storage
Compute
Display
Digital Video Delivery
GraphicsInterConnect
SAN/ NAS
Visualization Services Over LAN/WAN
Clients
Compute Cluster
7
Who Uses Grids
8
Life Sciences

• Genetic Sequencing
• Custom research or limited use applications
• Large Database searches and manipulations
• Exponential Combinations
• Limited administrative staff
• Complementary techniques

9
MCAE / MCAD

• Improve resource utilization
• Crash / Safety simulation (highly parallel)
• Computational Fluid Dynamics (highly parallel)
• Noise, Vibration, Harshness simulations
• Structural design / analysis
• Application license management

10
Electronic Design Automation

• Shorten time to market
• Increase chip design complexity
• Faster turn-around in verification / simulation
• Improve layout productivity
• Efficient EDA application license management
• Centralized control of design data
• Better Configuration Management (check in /
  check out)

11
Financial Services

• Risk analysis
• Portfolio Analysis
• Market simulations
• Time IS Money
• Proprietary applications
• Data mining
• Economic modeling
• Multiple scenario execution
• Need instant results analysis tools

12
Oil Gas, Geophysics

• Seismic Analysis
• Parallel Reservoir Simulations
• Geophysical Ray Tracing
• Visualizations
• Custom in-house codes
• Large scale, multi-platform execution

13
Asia Pacific Initiative

• This page is left blank

14
Model

• Technology to solve Complexity
• Innovative technology to decrease complexity
• Partnership
• Co-development of technology
• Help to jumpstart into the technology and market
• Partner to do the delivery
• Collaboration
• Joint research into key technology
• Universities, research institutes and partners
• Joint demonstration of technology

15
Sun Grid Engine on the ApGrid Testbed
622Mbps x 2
16
(No Transcript)
17
NTU-CampusGrid
18
Campus Grid Connectivity
19
www.ntu-cg.ntu.edu.sg
20
Design Grid Problem Solving Environment
This project vision provides an Internet portal
giving remote access to general analysis and
design tools. Our development road map for this
takes a route initially for Computational Fluid
Dynamic, Finite Element and Multi-disciplinary
Analysis and Design Grid Optimisation for a
diverse of science and engineering applications.
21
Portfolio Rebalancing in a Grid Computing
Framework

• We present an approach to compute the efficient
  frontier for portfolio optimization based on
  evolutionary programming (EP) technique. Our
  approach relies on multiple EP runs within a
  search to create the frontier.

22
USM/UKM Grid
23
GriDE Overview
GriDE is an integrated development environment
that make it straightforward for scientists and
engineers to construct grid applications. It
provides friendly tools to access grid resources
and makes the development approach easily and
fast. Portability Modular
Scalability Security Convenience
Transparency
24
GriDE Architecture
25
What is the BioBox?
26
Deployment Architecture of Bio-ClusterGrid
Homology and Similarity Search Biodas, BLAST,
FASTA, GlimmerM, Wise Sequence Analysis ACT,
ClusterW, EMBOSS, HMMER, Image,
T-Coffee Structural Prediction DOWSER,
FastDNAml, LOOPP, MapMaker/QTL, PAML,
PHYLIP Molecular Imaging/Modeling Artemis, Cn3D,
GROMACS, NAMD, NMRView, RasMol, ReadSeq,
TribeMCL, VMD Others Biojava, Bioperl, Biopython
Execution Tier (execution hosts) Executes jobs
sent to it by Grid Master Hosts Returns job
status/results to Grid Master Hosts
Resource Management/Portal Tier GEP provides
entry point for job submission SGE schedules jobs
to execution hosts SGE monitors job
status/completion of jobs GEP presents job
status/results to users SGE/GEP server acts as
file server
Solaris 9, Sun Grid Engine, Grid Engine
Portal Customer/ISV applications HPC Cluster
Tools, SunOne Studio 8 Compiler
Sunray Server 2.0 Or other desktop operating
systems
Solaris 9, Sun Grid Engine, Grid Engine
Portal Customer/ISV applications HPC Cluster
Tools, SunOne Studio 8 Compiler
27

• Easy-to-deploy installation package consisting of
  Sun OS and most popular Biox applications.
• Users who aren't familiar or want to avoid
  compilation/installation of OS/Biox applications.

28
Application Categories

• Homology Similarity Search (Blast)
• Sequence Analysis (Hmmer)
• Structural Prediction (Phylip)
• Molecular Imaging/Modeling(NAMD)
• Others (Biojava)

29
Submitting BLAST job using GEP
30
Other Reseach Collaboration

• GIS
• Digital Biologist
• SBS/NTU
• Bio/Med Grid
• USM/UKM
• Bio Grid
• BioTec (Thailand)
• Bio Cluster and BioX Applications
• ASIT (Philippines)
• Biobox initiative (training)
• University of Hokkiado
• KRIBB (Korea)
• Univ of Queensland (Aus)

31
Contacts

• Technical Contact
• Dr Simon See (simon.see_at_sun.com)
• APSTC (info_at_apstc.sun.com.sg)
• Website
• http//apstc.sun.com.sg
• http//www.sun.com/edu

32
The Productivity ChallengeImproving Resource
Access
Problem
Not using Scientists/Engineering/Creative
Technical professionals efficiently
Solution
A Grid makes it easy for the engineers to submit
jobs. They run more tests product design cycle
improves.
Benefits
Increase productivity which leads to shorter time
to market, higher quality and lower costs
33
Customer Challenges
Time-to-market pressures Multiple concurrent
projects with different schedules milestones
Growing datasets larger, more complex designs
In-house IT resource contraints Not enough
throughputmore and faster compute resources
required Heightened budget/ROI focus
34
Vertical Industry Challenges

• The need for more compute power, exponential
  increase in complexity
• Electronic Design Simulations, verifications,
  regression tests
• Automotive Crash tests, stress tests,
  aerodynamics, visualization
• Life Sciences Genetic sequencing, database
  queries
• Scientific Research Large compute needs,
  collaboration
• Oil and Gas Simulations, visualization, data
  analysis
• Digital Content Creation Frame rendering
• Software Development Code builds

35
The Grid Blueprint for a New Computing
Infrastructure
The grid is an emerging infrastructure that will
fundamentally change the way we think about and
use computing. The grid will... create a
universal source of computing power.
36
Operational Benefits of Grid Computing

• Improve utilization of computing resources and
  staff
• Fair, systematic way of workload management
• Tackle large tasks not possible before
• Effective way to deploy new infrastructure
  components incrementally
• Reduce operational complexity with lower
  management and control costs
• Product designers not acting as system admins

37
The Business Benefits

• Improved ROI on equipment, application licenses,
  staffing
• Shorter time-to-market
• Faster development and test cycles
• Increase productivity for people and machines
• A foundation for enterprise-wide collaboration

Each of our scientists has one extra hour every
day to work on other things. Duncan Ross,
Director of IT, Cognigen