SYSTEMATIC SDN AND NFV WORKSHOP 2014 BUSINESS AND ECONOMIC ASPECTS OF SDN AND NFV Bell Labs May 2014 1
BUSINESS AND ECONOMIC ASPECTS OF SDN AND NFV AGENDA 1. Introduction Challenging times for operators 2. Why embrace SDN/NFV? 3. Opportunity drivers for virtualization 4. Leveraging Cloud attributes 5. Monetization of SDN/NFV capabilities - illustration 6. Conclusion 2
CHALLENGING TIMES FOR OPERATORS +30% +32% +346% FIXED BROADBAND CONNECTIONS 532 Million 691 Million MOBILE CONNECTIONS 5.5 Billion 7.3 Billion WI-FI PUBLIC SPOTS 1.3 Million 5.8 Million +990% M2M DEVICES 1.1 Billion 12 Billion SERVICE PROVIDER COSTS AND PROFITS 70% Adjusted OPEX/revenue 60% 2011 2015 APPLICATIONS DOWNLOADS 18.2 Billion 41.7 Billion 2011 2015 SMARTPHONE SALES 428 Million 1 Billion 2011 +129% +134% +100% ENTERPRISE CLOUD SERVICES* 12 Billion 2015 25 Billion 2011 2015 +879% MOBILE VIDEO CONSUMPTION 429 Million 4.2 Billion // 20% 10% Adjusted operating profit/revenue CAPEX / revenue 2007 2008 2009 2010 2011 2011 2015 2011 2015 2011 2015 2011 2015 N E E D TO C O P E W I T H A G R O W I N G D E M A N D A N D C O S T S I N C R E A SE FA S T E R T H A N R E V E N U E S *Source: Yankee 3
WHY EMBRACE NFV AND SDN? DRIVERS FOR ADOPTION NFV SDN Scale services up and down quickly 90% Create new services quickly 81% Use standard virtualized servers to reduce costs 81% Optimize network configurations in real time 71% Introduce new services quickly 76% Simplify network provisioning 67% Optimize network in real time 67% Create new virtual networks quickly 67% Multi tenants on the same hardware 52% Create virtual networks across multivendor equipment 57% Reduce energy consumption 43% Use lower cost routers and switches 52% Introduce services based on geography or customer sets 38% Create new types of virtual networks 52% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Infonetics: SDN and NFV Strategies: Global Service Provider Survey ; July 8, 2013 O P E R AT IONA L E F F I C I E N CI ES, A G I L I T Y A N D N E W R E V E N U E O P P O RT U NI TIES M A K E V I RT U A L IZATION A P P E A L I NG 4
WHY EMBRACE NFV AND SDN? WHAT DOES APPLICATION VIRTUALIZATION BRING Traditional Application Platform Virtualized Application Platform Application on dedicated custom HW Application placed in a DC with other applications having complimentary usage profile Application on virtualized, high-volume, general-purpose HW in multi-tenant mode HW CAPEX SAVINGS APPLICATIONS SCALABILITY EARLY MOVER ADVANTAGE NEW BUSINESS MODELS % US$ Dedicated Servers Virtual Servers 100% 80% 60% 1 st Entrant 2 nd Entrant 3 rd Entrant 4 th Entrant 30% 20% 40% 20% 10% Baseline TBB 0.01M 0.5M 1M 5M 10M Subscribers 1 M 2 M 3 M 4 M 5 M Subscribers 0% 1-Player 2-Players 3-Players 4-Players Competition Scenario 0% 1 2 3 4 5 6 7 8 9 10 Year 60%-80% Savings Lower/Incremental Investment as Subscribers Increase Higher Market Share for Early Movers Further Share Gain from Try-Before-Buy (TBB) COMPELLING CASE DRIVEN BY BOTH REVENUE AND COST BENEFITS 5
WHY EMBRACE NFV AND SDN? VIRTUALIZATION ECONOMICS NETWORK SLICING ELASTIC SCALING RAPID DEPLOYMENT REVENUE MODEL TCO MODEL Virtualized General-Purpose Hardware CONCEPTS NETWORK SLICING NEW SERVICES NEW CUSTOMERS ENHANCED CONTROL DYNAMIC PRICING ELASTIC SCALING DESCRIPTION New & customized applications; Service Chaining Smaller customers, MVNOs & enterprises requiring specific network functionalities Offering business customers differentiated classes of service Time-, location- and context-sensitive pricing of services Offering existing services on a dynamic basis CAPEX Improved Asset Utilization Greater Scalability RAPID DEPLOYMENT Fast implementation of new services; Try before you buy offers Operations Automation and Standardization OPEX HW & SW Maintenance Reduction Energy Efficiency B O T H R E V E N U E G A I N S A N D C O S T S AV I N G S D R I V E V I RT U AL I ZATION E C O N O M I CS 6
OPPORTUNITY DRIVERS FOR VIRTUALIZATION VIRTUALIZATION NEEDS TO BE SELECTIVE IN THE NETWORKS YES LIKELY UNLIKELY *Dynamic Control enables new services Potential for new revenues with the right business model N E E D D E TA I L E D M O D E L S : V I RT U A L IZATION B E N E F I T S N O T O B V I O U S F O R A L L F U N C T I O NS 7
OPPORTUNITY DRIVERS FOR VIRTUALIZATION CAPTURING COST BENEFITS OF VIRTUALIZATION Value Drivers for TCO Gains APP ONBOARDING AUTO-SCALING Indicates Relative Progress SELF HEALING MULTI-TENANCY RESOURCE POOLING HW VIRTUALIZATION GENERAL PURPOSE HW DEDICATED NATIVE HW DEDICATED COTS HW VIRTUAL HW CARRIER CLOUD Technology Adoption Stage UNDERSTANDING TCO BENEFITS FOR VARIOUS SCENARIOS HELPS RATIONALIZE PRICING AND IMPROVE POSITIONING 8
LEVERAGING CLOUD ATTRIBUTES DRIVERS OF TCO SAVINGS M2M CASE STUDY Virtual Attribute Impact TCO Gain General-purpose processors Multi-tenancy Auto-scaling Lower-priced hardware Standardized support Higher asset utilization Smaller platform Elastic capacity provisioning HW CapEx reduction Platform services and Operations OpEx savings HW & SW CapEx reduction Platform services OpEx savings HW & SW CapEx reduction Platform services OpEx savings Self-healing Faster recovery from failures Operations OpEx savings Agility Software Standardization Scalability Resource Pooling Faster to market Rapid order provisioning Lower-priced software Standardized support Lower incremental investment per user Higher asset utilization Operations OpEx savings SW CapEx reduction Platform services and Operations OpEx savings HW & SW CapEx reduction Platform services OpEx savings HW & SW CapEx reduction Platform services OpEx savings 9
LEVERAGING CLOUD ATTRIBUTES M2M CASE STUDY - TCO SAVINGS SUMMARY $37M BASE SCENARIO 20M Devices 1 TCO -38% OpEx CapEx $23M 38% reduction in 5-yr TCO 37% CapEx savings 90% reduction through general-purpose HW use 32% increase in SW 15% HW savings due to the impact of cloud attributes: Multi-tenancy, auto-scaling, self-healing 40% OpEx reduction 24% reduction in lower maintenance & upgrade costs 58% reduction in power and real estate 46% reduction in operations cost NATIVE VIRTUAL 1 Number of devices in 2020 following 5-year growth SIGNIFICANT COST SAVINGS ARE REALIZED EVEN FOR SMALL DEVICE POPULATION 10
LEVERAGING CLOUD ATTRIBUTES IMPACT OF DEPLOYMENT SIZE ON TCO SAVINGS SCENARIO 2 Smaller Growth BASE SCENARIO Larger Growth SCENARIO 3 10M Devices 20M Devices 100M Devices $20.8M TCO -30% v $14.6M $36.5M TCO -38% v $22.6M $168.1M TCO -43% v $95.2M $11.3M CAPEX -30% v $7.9M $9.5M OPEX -29% v $6.7M $21.2M CapEx -37% v $13.4M $15.2M OpEx -40% v $9.2M CAPEX -40% $102.6MB v $61.6M $65.5M OPEX -49% v $33.6M VIRTUALIZATION GAINS ARE ROBUST WITH RESPECT TO DEPLOYMENT SIZE AND GROWTH 11
LEVERAGING CLOUD ATTRIBUTES APPLICATION SPECIFIC TCO SAVINGS SGW PGW MME PCRF TCO TCO TCO TCO $10.8M -53% OpEx CapEx $5.4M $8.0M -31% OpEx CapEx $5.7M $10.3M -42% OpEx CapEx $6.2M $6.7M -16% OpEx CapEx $5.9M NATIVE SGW VIRTUAL SGW NATIVE PGW VIRTUAL PGW NATIVE MME VIRTUAL MME NATIVE PCRF VIRTUAL PCRF GREATER COST GAINS ARE OBTAINED FOR APPLICATIONS ON SPECIALIZED PLATFORMS Values shown for Base Scenario 12
LEVERAGING CLOUD ATTRIBUTES MODELING OPERATIONS COST Performance Management - Native Monitor/Analyze Resource Performance Assign Resource Performance for Resolution Fault RCA and Resolution PO Issue HW Delivery to central site Obtain licenses Pre-assemble Ship to site Cabling, network config. S/P GW Integration Monitor/Analyze Resource Performance Identify/Resolve Fault Root-Cause Analysis Scale Site Survey Obtain licenses Pre-assemble On-board S/P GW deployment 0 20 40 60 80 100 120 140 160 Performance Management - Virtual Days Key Operations Related Contributors: Sources of failure for individual functions Time and effort currently required to perform: Issue identification Trigger & execute solution process Root cause analysis Time and effort required for each of these stages with virtualized functions Realizable savings in time and cost with virtualization 0 20 40 60 80 100 120 140 160 Days BOTTOMS UP OPERATIONS COST MODEL CAN BE DEVELOPED THROUGH A JOINT WORK PLAN WITH OPERATORS 13
LEVERAGING CLOUD ATTRIBUTES MODELING GAINS FROM MULTI-TENANCY Multi-Tenancy implies the execution of multiple network functions & applications on the same set of virtualized generalpurpose servers, resulting in sharing of HW resources. GROUPING OF CAPACITY REQUIREMENTS (ILLUSTRATIVE) 200% 180% 100% 160% 90% 140% 80% 70% 120% 60% 100% 50% 80% 40% 60% 30% 40% 20% CAPACITY CAPACITY REQUIREMENTS New Combined (ILLUSTRATIVE) Over provisioning level < X%+Y% Previous Combined Over provisioning level (X% + Y%) Average utilization Average utilization Over Over provisioninglevel= X% Y% Capacity savings Average utilization IPTV 10% 20% 0% 0% 1h 3h 5h 7h 9h 11h 13h 15h 17h 17h 19h 19h 21h 21h 23h 23h 1h 3h 5h 7h 9h 11h 13h 15h 17h 19h 21h 23h VDI VDI IPTV 43% 58% Variations in application demand have to be met by over-provisioning each application server individually. Pooling applications on general-purpose servers reduces variability of required server capacity and over-provisioning level. 41% INCREASE IN SERVER UTILIZATION LEADS TO CAPEX AND OPEX SAVINGS WITH NETWORK FUNCTION VIRTUALIZATION 14
MONETIZATION OF SDN/NFV CAPABILITIES CLOUDGAIN BELL LABS NEW REVENUE POTENTIAL ESTIMATOR TOOL PARAMETRIC SCENARIO DEFINITIONS Years 1 5 Adoption Start 1 1 10 Adoption Period 5 Impact Name Network Slicing High3 Elastic Scaling High 3 Faster TTM High [000s] 2014 2015 2016 2017 2018 Revenue TRUE View LTE Subscribers 67,578 91,703 113,800 116,645 119,561 Include Revenue TRUE Usage MB/Month 0% 100% 1X 10X 600 CAPEX Retention 100% 100 5yr Growth 5X Type Wireless Verizon Wireless Region Illustrative Results for NAR operator v Scenario (Monetizing Dynamic Control) $18 Billion new revenues over 5 years Network Slicing is the biggest driver Share of new revenues can reach 12% in Year 5 NAR Existing Customers Incremental Revenue Matrix [$M] Existing New Total Services Services 7,659 8,310 15,969 NEW REVENUE STREAMS Dynamic Control Gain [Revenue $M] 4,082 1,716 12,641 Network Slicing Components [$M] Dynamic Pricing Enhanced New Network Slicing Elastic Scaling Shorter TTM Control 2,037 Customers 433 2,470 New Customers Total 9,696 8,743 18,439 Dynamic Control Gain [%Revenue] New % Value = 6.3% Services 14% 12% 0 2,000 4,000 6,000 10% 8% 6% 4% 2% 0% 2014 2015 2016 2017 2018 THE TOOL WILL BE ENHANCED TO COVER MORE VALUE DRIVERS AND USE 15
MONETIZATION OF SDN/NFV CAPABILITIES CONNECTED CAR USE CASE Mobile Access MME SGW HSS PGW DC 1 DC 2 Mobile Core Internet BW Charges Current QoS Operator Enhanced QoS New Revenues Enterprise HQ AUTO MANUFACTURER APPLICATION One Car, Multiple Services CURRENT SERVICES CoS LEVEL ENHANCED SERVICES CoS LEVEL Current Services Enhanced Services NAVIGATION BRONZE GOLD SAFETY BRONZE GOLD Service Fee CONSUMER Incremental Fee TELEMATICS BRONZE SILVER INFOTAINMENT BRONZE SILVER INTERNET ACCESS BRONZE BRONZE 16
MONETIZATION OF SDN/NFV CAPABILITIES CONNECTED CAR USE CASE NEW REVENUES Revenue 5% Elastic Scaling % BW Increase Revenue $100M $80M Rapid Deployment Revenue 600 MB Enhanced Control BW Increase $60M 4% 3% 2% 1% $80M $60M $40M $20M $60M + + $40M $20M 500 MB 400 MB 300 MB 200 MB 100 MB $50M $40M $30M $20M $10M 0% 2014 2015 2016 2017 2018 2019 2020 $0M $0M 2014 2015 2016 2017 2018 2019 2020 0 MB 2014 2015 2016 2017 2018 2019 2020 $0M $200M New Revenues $160M $120M $80M Elastic Scaling Rapid Deployment Enhanced Control 7-year New Revenue: > $600 Million $40M $0M 2014 2015 2016 2017 2018 2019 2020 30% Revenue Gain MONETIZING DYNAMIC CONTROL, AGILITY AND SERVICE GRANULARITY 17
CONCLUSION SDN/NFV can deliver on its promises for addressing the challenges but it is NOT a magic wand SDN/NFV has a strong impact on network operations and a lot of traditional rules have to be revisited Service agility and new business opportunities should be considered in addition to the cost benefits Only implementation of SDN/NFV performed on a large scale will allow to face the market conditions in the long run 18
CONTACT Olivier Kermin ALCATEL-LUCENT BELL LABS BUSINESS MODELING Tel : +33 (0) 1 60 40 31 61 Cell : +33 (0) 6 80 47 81 55 olivier.kermin@bell-labs.com 19