IBM Symposium 2004

1
What is Cloud Computing? Bowman Hall Program
Director Cloud Computing Client Engagements IBM
Software Group bow_at_us.ibm.com
2
IT infrastructure is reaching a breaking point.
85 idle
1.5x
70 per 1
In distributed computing environments, up to 85
of computing capacity sits idle.
Explosion of information driving 54 growth in
storage shipments every year.
70 on average is spent on maintaining current IT
infrastructures versus adding new capabilities.
40 billion
33
Consumer product and retail industries lose about
40 billion annually, or 3.5 percent of their
sales, due to supply chain inefficiencies.
33 of consumers notified of a security breach
will terminate their relationship with the
company they perceive as responsible.
3
As the world gets smarter, demands on IT will
grow
Intelligent oil field technologies
Smart traffic systems
Smart energy grids
Smart healthcare
Smart food systems
Smart retail
Smart water management
Smart regions
Smart weather
Smart countries
Smart supply chains
Smart cities
4
Infrastructure needs to become more dynamic
Business and IT Workloads
Facilities Infrastructure
Production Infrastructure
MobilityInfrastructure
TechnologyInfrastructure
Communications Infrastructure




to bring together business and IT to create
new possibilities.
5
A Crisis of Complexity - The Need for Progress is
Clear.
Global Annual Server Spending (IDC)
300
Power and cooling costs
Management and admin costs
250
New system spend
200
Uncontrolled management and energy costs
150
100
50
Steady CAPEX spend
0B
1997
1996
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
To make progress, delivery organizations must
address the server, storage and network operating
cost problem, not just CAPEX
Source IBM Corporate Strategy analysis of IDC
data
6
Cloud Workload Optimized Consumption Delivery
Models

• Cloud is a new consumption and delivery model
  inspired by consumer Internet services.
• Cloud enables
• Self-service
• Sourcing options
• Economies-of-scale
• Cloud represents
• The Industrialization of Delivery for IT
  supported Services
• Multiple Types of Clouds will co-exist
• Private, Public and Hybrid
• Workload and / or Programming Model Specific

Cloud Services
Cloud Computing Model
7
Cloud Computing Delivery Models
Collaboration
CRM/ERP/HR
Business Processes
Industry Applications
Software as a Service
Middleware
Web 2.0 Application Runtime
Java Runtime
High Volume Transactions
Development Tooling
Database
Platform as a Service
Servers
Networking
Storage
Data Center Fabric
Shared virtualized, dynamic provisioning
Infrastructure as a Service
8
Cloud Computing Deployment Models
Flexible Delivery Models

• Public
• Service provider owned and managed.
• Access by subscription
• Delivers select set of standardized business
  process, application and/or infrastructure
  services on a flexible price per use basis.

• Private
• Privately owned and managed.
• Access limited to client and its partner network.
• Drives efficiency, standardization and best
  practices while retaining greater customization
  and control

Cloud Services
Cloud Computing Model

• Hybrid
• Access to client, partner network, and third
  party resources

. Customization, efficiency, availability,
resiliency, security and privacy
.Standardization, capital preservation,
flexibility and time to deploy
ORGANIZATION CULTURE GOVERNANCE
...service sourcing and service value
9
A stepwise approach to Cloud Computing

• Create an IT Transformation Roadmap
• Define an Architectural Model for Cloud Computing
• Complete a Workload Analysis
• Decide the Right Mix of Delivery Models
• Implement a Cloud Solution

10
Step 1 IT Transformation Roadmap

• Standardized Services
• Dramatically reduce deployment cycles
• Granular service metering and billing
• Massively scalable
• Autonomic
• Flexible delivery enables new processes and
  services

• Remove physical resource boundaries
• Increased hardware utilization
• Allocate less than physical boundary
• Reduce hardware costs
• Simplify deployments

• Reduce infrastructure complexity
• Reduce staffing requirements
• Improve business resilience (manage fewer
  things better)
• Improve operational costs/reduce TCO

Dynamic
Automate
Shared
Virtualize
Simplified
Consolidate
11
Step 2 Architectural Model for Cloud Computing
Service Request Operations
Service Provider
Service Creation Deployment
End Users,Operators
ServicePlanning
Cloud Services
Service Definition Tools
Service Publishing Tools
Role-basedAccess
Service Fulfillment Configuration Tools
Cloud Management Platform
Service Catalog
Service Reporting Analytics
Operational Console
12
Step 3 Workload Analysis
Higher Gain From External Cloud
Collaboration
SME ERP/SCM/CRM
Numerical Low Data/Compute
Web Serving
Data Warehousing
Start Here
Data Mining
Numerical High Data Transfer
Virtual Desktop
Higher Pain To Cloud Delivery
Application Devt. Test
Systems Mgmt.
File Print
Lower Pain To Cloud Delivery
LE - ERP/SCM/CRM
LE - Transaction Processing
Content Centric Architecture
Storage - Analytics Architecture
Lower Gain From External Cloud
13
Step 4 Deciding the Right Mix of Delivery Models
Managed Operations
Off Premises Shared
Public Cloud Services
Off Premises Dedicated
Delivery Models
On Premises Utility
Private Cloud Services
Traditional IT
On Premises
Fixed
Variable
Mixed
Financial Models
14
Step 5 Self-service drives Process
Standardization
End Users
Service Portal
Service Request Catalog

• Provisioning Engine
• Workflows
• Expert Systems
• Scripts

• Optional Service Modules
• e.g. Metering/ Usage Billing, Monitoring, etc.

• Benefits
• Lower cost
• Ease-of-use and access
• Process transformation

Virtualized Cloud Infrastructure
15
Traditional Data Center Management
vs.Cloud-like Management
For Cloud-like efficiencies and flexibility, it
is not sufficient to have the right technology,
but to also use it in the right way!
The overall objective of Cloud-managed data
centers is to automate any type oftask or
situation (by reducing manual intervention) for
increasing flexibility and reducing operational
expenses
Days/weeks
11000s
efficiency
Hours/minutes/sec
flexibility
150-1100
Cloud
Traditional mgmt
16
IBM Cloud Computing Products Services
Development and Test
Infrastructure (compute / storage)
Desktop and Devices
BusinessServices
Analytics
Collaboration
Smart Business on the IBM Cloud
Smart Business Desktop on the IBM Cloud
Information Protection Services Computing on
Demand
Standardized services on the IBM Cloud
Smart Business Development Test on the IBM Cloud
Lotus Live
Smart Business Cloud
Private cloud services, behind your firewall,
built and/or managed by IBM
Smart Business Desktop Cloud
Scale out File Services
Smart Business Test Cloud
Smart Business Systems
Pre-integrated, workload optimized systems
Smart Business for SMB (backed by the IBM cloud)
Smart Analytics System Powered by Infosphere
IBM CloudBurst w/Quickstart Svcs
IBM CloudBurst w/Quickstart Svces
Future
Available
17
IBMs approach is based on our own transformation
IBM IT Transformation

• From 2002 through 2007, IBM's own IT investments
  delivered a cumulative benefit yield of
  approximately 4 billion. For every dollar
  invested, we saw a 4 cumulative benefit.

• Consolidation and virtualization - thousands of
  servers onto approximately 30 IBM System z
  mainframes.
• Additional virtualization leveraging System p,
  System x and storage across enterprise.
• Substantial savings being achieved in multiple
  dimensions energy, software and system
  management and support costs.

Data Center Efficiencies Achieved

• The virtualized environment will use 80 less
  energy and 85 less floor space.
• 2X existing capacity, no increase in consumption
  or impact by 2010.

Project Big Green
Cloud-enabled on demand IT delivery solution

• Self-service for 3,000 IBM researchers across 8
  countries.
• Real time integration of information and business
  services.

18
IBM Technical Adoption Program (TAP)ROI Analysis
Reduced Capital Expenditure Reduced
Operations Expenditure Additional
Benefits Reduced risk, less idle time, more
efficient use of energy, acceleration of
innovation projects, enhanced customer service
Without Cloud
With Cloud
100
Strategic Change Capacity
Software Costs
Power Costs
Current IT Spend
Labor Costs (Operations and Maintenance)
Business Case Results Annual savings 3.3M
(84) from 3.9M to 0.6M Payback Period 73
days Net Present Value (NPV) 7.5M Internal
Rate of Return (IRR) 496 Return On Investment
(ROI) 1039
Software Costs
Hardware, labor power savings reduced annual
cost of operation by 84
Power Costs (- 89)
19
In Summary

• Enterprise cloud opportunity is early, but real
• Economics of private clouds look compelling
• Adoption of cloud computing will be driven by
  workload affinity
• IBM offers deployment choices for workloads that
  matter to you