1. Smarter Computing
Understand the likely impact of Stream
Computing and Big Data on your Business
and Data Center
How to control costs, improve performance, and mitigate risk
throughout the data lifecycle
2. IT needs to reinvent itself in order to enter this new era of
computing
Entering New Era of Computing
Driven by…
• Insatiable
Demand
Centralized Distributed Smarter
Computing Computing • Unsustainable Computing
Economics
• New
Technologies
Efficient Innovative Efficient &
Innovative
but … but …
Limited access Fueled sprawl &
& flexibility increased TCO
1
3. The information supply chain
Transactional &
Collaborative Analyze Business Analytics
Applications Applications
Content
Integrate Analytics
Big Data
Master
Data
Manage Cubes
Streams
Data
Data
External Content Warehouses
Information
Sources Streaming
Information
Govern
Security &
Quality Lifecycle Standards
Privacy
2
2
4. What is ‘BIG’ Data
4 Billion 10x
# of cell phone users worldwide Growth in digital data every 5 years
2 Billion 5%
# of Internet users worldwide New Information that is structured
“ When physical assets such as cell phones, traffic sensors, cameras, PCs, RFID tags etc.
become elements of an information system with ability to capture, compute and communicate
information themselves on a massive sensor scale using common TCP/IP protocol.
-- Internet of Things by McKinsey Global Institute, 2010
“
The Industrial Revolution of Data.
-- Dr. Joe Hellerstein, UC Berkeley
3
5. The opportunity for real-time analytic processing
is everywhere …
Stock market
• Impact of weather on
securities prices
• Analyze market data at Law Enforcement,
ultra-low latencies
Defense & Cyber Security
Natural Systems • Real-time multimodal surveillance
• Wildfire management • Situational awareness
• Water management • Cyber security detection
Transportation Fraud prevention
• Intelligent traffic • Detecting multi-party fraud
management • Real time fraud prevention
Manufacturing e-Science
• Process control for • Space weather prediction
microchip fabrication • Detection of transient events
• Synchrotron atomic research
Health & Life Sciences
• Neonatal ICU monitoring Other
• Epidemic early warning Telephony • Smart Grid
system • CDR processing • Text Analysis
• Social analysis • Who’s Talking to Whom?
• Remote healthcare
monitoring • Churn prediction • ERP for Commodities
• Geomapping • FPGA Acceleration
4
6. Evolution of technology frontiers
Dat
a in
M ot
ion
Real Time Analytic Processing
(RTAP) to improve business
Data at rest response
Analysis of historic data
to improve business Stream
Reporting and transactions Computing
human analysis on
historical data Data
Warehousing 2003
Operational
Databases 1983
1970 DB2 v1
1968 Relational
database
Hierarchical
database
OLTP OLAP RTAP
Online Transaction Processing Online Analytical Processing ‘Real-Time’ Analytical Processing
6
7. The Big Data Challenge
• Manage and benefit from massive and growing amounts of data
• Handle uncertainty around format variability and velocity of data
• Handle unstructured data
• Exploit BIG Data in a timely and cost effective fashion
COLLECT MANAGE
Collect Manage
Integrate
INTEGRATE Analyze
ANALYZE
7
8. IBM offers a comprehensive and integrated set of solutions
for many types of BIG Data processing needs
Str e
ams
BigInsights filte
Non-Traditional rs in
com
i ng d
Data ata
ere
WAREHOUSE foSph lytic
s In S Ana
Traditional use S
s re or SP
m se
Data a
Stre ehou
r
Wa els
d
mo
Persistent In-Motion
Data Data
9
9. The Big Data ecosystem: Interoperability is key
BigInsights
User results
WAREHOUSE
Data Sources
Data can flow into one or
more of the environments
(even simultaneously)
10
10. IBM InfoSphere Streams
Stream RTAP
Analytics
In-Motion
Ultra Low
Non- Analytics
Latency Results
Traditional /
Non- Relational
Data Sources
OLAP / OLTP
Traditional
Analytics
Traditional /
Relational Data
Sources
At-Rest
Database Results
Data
Analytics
11
11. InfoSphere Streams for companies who need to …
Real-time delivery
ICU Environment
• Deal with Gigabytes of data each Monitoring Monitoring
second Algo Powerful Telco churn
Trading Analytics predict
• Work with application, sensor Cyber Smart
Security Government / Grid
and internet data, video/audio Law enforcement
• Deliver insight in microseconds
to analytical applications
Millions of
events per Microsecond
• Support complex scenarios using Latency
second
C++ or Java code
• Integrate with existing analytics
Traditional /
& data warehousing investments Non-traditional
data sources
12
12. Cyber security example – Detecting botnets
Real-time Data-at-rest
Analytics Analytics
BigInsights
Custom
Retain Visualizations
Aggregate
+ Results
Real-time
Window
Real-time Data
(Traffic Capture,
Alerts, Logs, …) Data-At-Rest Window
Aggregated TRAFFIC + DNS + Secondary Information
Alerts etc.
PCAP/
Now Aggregated Past PCAP/
Netflow
Netflow
Alerts
IBM Commercially Sensitive
15
13. IBM is uniquely positioned to help organizations handle their
“BIG Data” analysis and management challenges
Integrate Automate Secure
• Scale to petabytes and thousands of users
• Deep integration with Cognos and SPSS
• Integrated analysis and analytic model consistency
17