SDxCentral SDN NFV Market Size Report 2015 A

Market Report

SDxCentral SDN and NFV Market Size Report 2015 Edition

The Trusted News and Resource Site for SDx, SDN, NFV, Cloud and Virtualization Infrastructure

Market Report

SDN and NFV Market Size Report 2015

Table of Contents Executive Summary..................................................................................................................................................................... 1 Examining the Data..................................................................................................................................................................... 1 Forecasting the SDN and NFV Markets............................................................................................................................ 2 Beginning with Assumptions................................................................................................................................................. 3 Baselining SDxN Adoption..................................................................................................................................................... 4 Reviewing the Case for an Updraft..................................................................................................................................... 4 The Case for a Downdraft....................................................................................................................................................... 5 Quantifying the SDN and NFV Impact.............................................................................................................................. 7 Mapping Impact to Customer Segments......................................................................................................................... 9 Translating Customers to Use Cases.................................................................................................................................. 11 The Quantitive Impacts of SDN and NFV........................................................................................................................ 12 Conclusion...................................................................................................................................................................................... 13

© 2015 SDNCentral. All Rights Reserved.

Page ii

Market Report

SDN and NFV Market Size Report 2015 Executive Summary The adoption of software-defined networking (SDN), network function virtualization (NFV), network virtualization (NV), and white box (gray box and brite box) initiatives has dramatically changed the networking landscape over the past few years. In response to the demand for data around the size and scope of the market, SDxCentral has updated its 2013 Market Sizing Report to reflect the realities of today. This 2015 Market Sizing Report not only provides market data, but also includes a comprehensive analysis of the SDx Networking trends, influences and use cases that are impacting uptake. Our updated model predicts the combined revenue of SDN, NFV and other next-generation networking initiatives (SDx Networking) will exceed $105B per annum by 2020. The rapid emergence of SDxN as an influencer in network purchasing decisions will dramatically impact what has been a largely stagnant network vendor competitive landscape. It’s expected these technologies will influence almost 80% of the purchasing decisions associated with all networking revenue by the end of 2020, affecting virtually every customer segment within the networking space. New SDxN use cases could open the market to new entrants for the first time in more than a decade. Combined with the trend towards more converged IT organizations, SDx Networking could enable a new type of networking decision maker who has less experience with and weaker ties to incumbent vendors. Cloud providers have been leading the charge into the new network paradigm and are expected to remain the largest consumer of new network technologies through 2020.

Examining the Data The SDxN market-sizing model is based on the broader networking Total Addressable Market (TAM). We are expecting the networking TAM to be approximately $100B in 2015, and grow to an estimated $132B by 2020. The TAM is projected to have a compounded annual growth rate (“CAGR”) of 5.8% over the next 5 years, from 2015 to 2020. This overall networking TAM was determined by looking at the following constituent elements: • Optical (MSPP/SDH, Optical Switch/OXC/NG Core Switch, WDM including POTP, MWDM, LHDWDM) • Wireline (FTTx access, copper line access, CMTS access, DSL modem & home gateways), cable CPE) • Core (IP core routers) • Edge Aggregation (IP edge routers, carrier edge Ethernet) • Wireless Access (3G & LTE small cells, CSP hot spots) • Enterprise (traditional routing, Ethernet switching, WLAN) • Network Functions (Firewall/VPN, IDS/IPS, integrated threat management, network access control, SSL VPN gateway, security information event management, specialized threat analysis & protection, data leak prevention, content security gateways, WAN Optimization Controllers, Application Delivery Controllers, SP deep packet inspection, mobile packet core, transparent caching, DDI, Sessions Border Controllers) • NMS (Network Management Systems – primarily made up of policy elements)


Page 1

Market Report


NETWORK MARKET TAM $140

$120

Revenue Value ($ Billions)

$100

$80

$60

$40

$20

$0 2015

2016

Optical

2017

Network (L2/L3)

2018

2019

Network Functions (L4-7)

2020 NMS

The base numbers are derived in part from publicly available data on market sizes and estimated market growth. Where there were multiple sources, data was blended using additional market research conducted by SDxCentral’s researcher. Where there was incomplete or no data available, SDxCentral research filled the gaps.

Forecasting the SDN and NFV Markets The objective of forecasting the size of the SDxN market is to estimate the amount of network spend that will be directly influenced by these new technologies. To put it simply, the goal is to ascertain when SDN, NFV, network virtualization or white/gray/brite boxes are a meaningful part of the selection/decision criterion (either for immediate or for future deployment). This forecasting is not intended to predict the rate of SDxN deployments in production environments, though it is a reasonable follow-on conclusion that these technologies will become a part of virtualized network deployment plans in the future. And therefore for our marketing sizing methodology, because we model the SDN and NFV impact through purchase criterion, our estimates will trend higher than actual SDN and NFV deployments. Regardless, we have seen that customers are considering how any networking solution they procure today will operate in an SDxN environment over the life of that equipment. In the case of software, buyers are examining the committed roadmaps and extensibility of solutions to ensure they can integrate with future virtualized environments.


Page 2

Market Report

SDN and NFV Market Size Report 2015 Essentially this SDN and NFV forecast is based on (a) product architectures that are shifting to central control and use programmatic APIs for general policy or flow-based control over data,

ENJOYING THE INSIGHTS OF THE REPORT SO FAR? CLICK TO SHARE WITH OTHERS!

(b) virtualized architectures that provide for easy deployment of network functions on commodity servers in a variety of packages. Note: some leading networking vendors are already selling their network function solutions by capacity, without regard for packaging and platform. For example, a buyer can purchase 100Gbps of IPS and deploy on a blend of VMs, bare metal or proprietary vendor-specific hardware. (c) commodity servers and switches with proprietary or open-source networking stacks (L2-4) that replace existing proprietary HW solutions. It is important to recognize that when sizing the SDxN market, we did not use OpenFlow-enabled Ethernet ports to determine the SDN opportunity; doing so would have ignored many relevant protocols, technologies and products that do not support OpenFlow, but are equally important to consider.

Beginning with Assumptions This SDx Networking market forecast is predicated on the assumption that: • SDxN related spend is largely substitutive to an existing network spend. • Different product types will be impacted by SDxN in different ways. • The uptake of SDxN use cases will dictate how much and when spending will shift from traditional to SDxN solutions. It is unlikely that SDxN solutions will increase network spend on its own; network spend tends to be tightly correlated with IT spend, which, at a macro level, is more closely tied to GDP than individual technologies. This means any SDxN spending must come at the expense of either a non-SDxN equipment spend or in place of operational savings that allow for additional purchases. Furthermore, the rise of SDxN will result in greater adoption of open-source solutions and commoditized hardware, which we believe will add to the erosion of overall revenues. We believe this erosion may be balanced by more networking purchases, which are needed to support the larger bandwidth requirements associated with video and the host of next generation apps that place a lot of demands on the network. The impact of SDxN cannot be modeled as a flat percentage applied equally to all product types in all market segments. Some products and markets will lend themselves to SDxN use cases more than others, and some buyers will be more or less predisposed to adopt new technologies. Ultimately, the specific SDxN use cases will dictate when different parts of the market will move towards SDN and NFV and to what extent they will make the transition. Those use cases that have the most potential will drive technology development and determine where SDxN spend is focused first.


Page 3

Market Report

SDN and NFV Market Size Report 2015 Baselining SDxN Adoption As in our previous report, we stand behind our use of VMware’s uptake in the overall server virtualization market as a reasonable proxy for adoption timeframes. Since VMware essentially created the market and had a near 100% market share, it is an excellent model to base our uptake of SDxN technologies. The server virtualization market grew from nothing to approximately $1B annually in just over six years and then increased more than 5 times over the next six years to exceed $6B annually. Using this baseline, it is reasonable to assume the SDx Networking market should materialize over roughly the same time period.

Reviewing the Case for an Updraft As indicated in the 2013 report, SDxN adoption is unlikely to follow the exact market growth curve of the server virtualization market. There are strong technology and market dynamics that could significantly accelerate SDxN penetration. An updated list of driving forces include: • Compute virtualization adoption serving as a catalyst for SDN and NFV adoption • Cloud computing adoption and IT convergence • Aggressive carrier efforts to monetize their network and reduce their capital and operational expenses

Compute Virtualization as a Catalyst Since VMware launched, customers have been educated on the capabilities and benefits of compute and server virtualization, adopted the solutions and catapulted the market. Ultimately, server virtualization has taken root and changed how both vendors and customers view the server market. This transition, because it was both stark and recent, serves as a strong case study for the impacts of massive technology shifts. As SDxN follows a similar trajectory, it seems unlikely the major networking incumbents will be complacent in making the transition to next-generation technologies; in fact, every major vendor has announced plans to support NFV and SDN, which could accelerate the maturation of the SDxN market. In addition, the fact that customers have already deployed virtualized solutions may also accelerate the adoption of networking technologies that can fit and better support those virtualized storage and server environments.

Cloud Computing and IT Convergence The adoption of cloud computing has been accelerating over the past several years, driven by its ability to provide organizations pools of public and private resources they can use to better scale, support and orchestrate their workloads. Cloud computing untethers the applications and underlying infrastructure, providing a level of dynamism that simply did not exist in traditional networks. It is driving a convergence in IT, enabling organizations to manage their infrastructure more holistically and support more unified share services. As cloud initiatives grow, so does the pressure on the network to match the flexibility, programmability and scale of these more virtualized environments. We have already started to see telco providers learn from and adopt similar architectures to their cloud provider peers to host network functions. To achieve a more flexible transport that can adjust in real-time to the needs of users and applications, providers are and will continue to turn to SDxN solutions.


Page 4

Market Report

SDN and NFV Market Size Report 2015 For example, a major draw of SDN is its ability to orchestrate the network centrally and deliver programmatic interfaces that enable applications to provide inputs into the underlying network. This bridging of the network-application divide is particularly attractive to organizations looking to converge their IT and remove boundaries to better support the applications that ultimately drive the business. Ultimately it is this confluence of cloud adoption and IT convergence that will provide ample use cases for SDxN technologies and drive ongoing demand.

Aggressive Carrier Efforts Historically, carriers have been hesitant to embrace anything that may disrupt the network, which is ultimately their business. They have been slow to change and innovate, however, this is no longer the case, as they see possibilities to reduce their CAPEX and better monetize their network with SDxN technologies. They have seen how the adoption of server virtualization and cloud computing has created a radical shift in the way compute resources are purchased, deployed, managed and scaled and are looking to derive the same benefits from their network. In addition to quickly adopting existing virtualization technologies, carriers have been aggressively driving network innovation through a variety of channels. In 2013, the Industry Specification Group (ISG) for NFV, within the European Telecommunications Standards Institute (ETSI), was created to accelerate the development of standards for virtualized network functions in a telecommunication’s network. It has been quickly putting forward draft standards that vendors have been developing too in equally rapid fashion. To make it easier for the industry to mobilize around new NFV products and services, the Open Platform for NFV Project (OPNFV) was announced (in September 2014. Carriers have been eager to trial the technologies that have come out of the collective efforts of the industry and open source community. They have been trialing solutions from the OpenStack/KVM ecosystem, OpenDaylight, ONOS and OPNFV projects. Early POCs have been focused on flushing out hidden issues that arise when working across vendors and carriers. The goal is to get SDN, NFV and other next-generation networking technologies production ready as fast as possible, so carriers can improve the flexibility, programmability and efficiency of their networks to improve their margins and profitability.

The Case for a Downdraft In spite of the increased number of people rallying around SDxN architectures, there are still elements of the technologies that can slow adoption, including the: • Complexity of the solutions • Shortage of adequate skill sets • Difficulty in integrating with real-world production systems • Inadequacy of virtualization infrastructures Some of these elements are new and others were present in our 2013 report. However, one of the original barriers to adoption—incumbent vendor behavior—is no longer considered a major impediment to adoption. Instead, we have seen both new and incumbent networking vendors use SDxN as a way to drive new sales and to attack each other’s market shares. In fact, the extent to which some of the major vendors have embraced these changes and invested in open-source efforts has been surprisingly positive. Nevertheless, there are still significant challenges to adoption of SDxN.


Page 5

Market Report

SDN and NFV Market Size Report 2015 Complexity Since 2013, we have seen many more SDN deployments in production. However, the rate of full SDN deployment is still a small fraction of what some of the more bullish pundits predicted. As we observed in 2013, the very nature of the network is to act as connective tissue between compute, storage, and services. Accordingly, the reach of SDN is significantly greater and the complexity of rolling out is much higher. On the NFV front, we are seeing similar challenges. Even though NFV deployments tend to be less complex in early proofs-of-concept, the lack of standardized orchestration and management suites get in the way as network operators try to scale up their deployments. As a means of ensuring investment protection, customers will likely demand SDN and NFV support even if they are months or years from deploying it in production.

Skill Sets As was true two years ago, a lack of training and the right skill-set within an organization is still an issue. With SDxN, the drive toward automated, highly programmable network behavior represents a significant shift in skill sets away from the highly specialized network engineer. As companies move to a more programming-savvy workforce, there will be a requirement to either train existing employees, or replace some network engineers with software developers. Retooling the workforce could slow SDN and NFV penetration.

Difficulty Integrating with Real-World Production Systems Some of the largest challenges with SDN and NFV are not in how to bring the technology to market in greenfield installations, but how to seamlessly integrate these new technologies with existing brownfield situations. In particular, for enterprises, the question of how SDN and NFV integrate with their orchestration systems remains and for service providers, there are still many open questions on how NFV integrates into existing OSS/BSS workflow. Furthermore, in SDN deployments, the awkwardness of trying to integrate SDN into existing networks or even interoperability between SDN and non-SDN networks remains trying, even for the most advanced network operators.

Inadequacy of Virtualization Infrastructure Particularly for NFV, some of the early deployments were besieged with performance issues. Most virtualization infrastructure focuses simply on isolation of boundaries between virtual machines, and does not provide advanced resource management to ensure that compute, I/O and storage resources are apportioned fairly. The issue with the noisy neighbor problem—where a co-located VM takes up all resources, leaving none for a critical network function— has not been solved. Vendors today continue to work on managing the predictability of performance across NFV infrastructure, without which network operators cannot feel comfortable migrating from reliable, predictable proprietary hardware to NFV solutions.

CLICK TO SHARE THIS REPORT WITH OTHERS!


Page 6

Market Report

SDN and NFV Market Size Report 2015 Quantifying the SDN and NFV Impact Assuming that SDxN spending is largely substitutive for that of existing networking solutions, the general networking TAM provides the overall market from which SDxN will pull. As we’ve discussed before, the server virtualization market provides a timeframe in which the SDxN market is likely to emerge.

SDX NETWORKING MARKET FORECAST BY PRODUCT CATEGORY $120


$100

$80

$60

$40

$20

$0 2015

2016

Optical

2017

Network (L2/L3)

2018

2019

Network Functions (L4-7)

2020 NMS

Our model shows potential SDxN revenues growing from less than $15B in 2015 to nearly $105B by 2020. This growth is driven primarily by the network (L2/3) and network functions (L4/7) categories. The L2/3 contribution is expected to exceed $10B in 2015 and grow to more than $105B by 2020, while the network functions layer will grow from under $5B in 2015 to $32B by 2020. The SDxN market is projected to have a CAGR of 44% over the 5 years from 2015 to 2020, eight times the growth rate of the broader TAM. The portion of network purchases influenced by virtualization is anticipated to increase from 16% in 2015 to almost 80% by 2020.


Page 7

Market Report


PORTION OF NETWORK PURCHASES INFLUENCED BY SDX NETWORKING 100% 90% 80%

Percent of TAM

70% 60% 50% 40% 30% 20% 20% 10% 0% 2015

2016

2017 SDx Networks

2018

2019

2020

Traditional Networks

In addition, a major portion of L2/L3 spend will migrate from HW spend to a new suite of software-only networking applications. We estimate that by 2020, the market for just L2-3 networking SW apps will be $14B. These will appear as applications running on controllers or integrated into provisioning or orchestration systems, or in some situations run partially as VNFs on NFV infrastructure (e.g. virtual route reflectors).

ENJOYING THE INSIGHTS OF THE REPORT SO FAR? CLICK TO SHARE WITH OTHERS!


Page 8

Market Report


SHIFT FROM TRADITIONAL L2-3 HARDWARE TO SOFTWARE 14000

12000

Revenue Value ($ Millions)

10000

8000

6000

4000

2000

$0 2015

2016

2017

2018

2019

2020

L2-3 Networking SW Apps

Mapping Impact to Customer Segments Having quantified the impact of SDN and NFV, we will next take two different views into this market forecast. We’ll start by mapping the product categories we described at the beginning of this report to what is used in each of the following customer segments and sub segments: • Enterprise • Global 500 • Very Large (>10000 employees) • Large (>1000 employees) • Medium (>100 employees) • Small (<100 employees) • Public Sector • Telco • Wireline • Wireless • Managed Service Provider


Page 9

Market Report

SDN and NFV Market Size Report 2015 • Cloud and SAAS • Hyperscale • Iaas & PaaS Providers • SaaS Providers • Cloud Hosting • Managed Services - Cloud We note that the impact of SDxN virtualization in these customer segments will not be uniformly felt. By running our model which estimates the level of uptake per segment, we arrive at the following breakdown:

SDX NETWORKING SPEND FORECAST BY CUSTOMER SEGMENT

$120


$100

$80

$60

$40

$20

$0 2015

2016 Cloud

2017

2018 Enterprise

2019

2020

Telco

Annual SDxN purchases for cloud providers is expected to grow from $7B in 2015 to nearly $43B by 2020, and the cloud providers are already a significant purchaser of SDxN technologies and will continue to be a big part of the overall market. Enterprises will lag Cloud providers slightly (spending $4B in 2015 and $37B in 2020) but as private cloud architectures take off and they mimic the Cloud providers’ architectures, SDxN spending will increase as well. And as for the telecom providers (Telco), we expect a decent uptick in growth, with a 46% CAGR, as they aggressively adopt NFV architectures and demand NFV and SDN capabilities of their networking vendors. For the Telcos, we anticipate them growing from just shy of $4B in 2015 to almost $26B in 2020. Even the smallest market segments will see a sizable increase in SDxN spend by the year 2020. The small and medium business segments, accounting for about $300M in SDxN spend in 2015, will each reach $2B annually by 2020.


Page 10

Market Report

SDN and NFV Market Size Report 2015 Translating Customers to Use Cases For this year’s report, we decided to break down the spending by use cases. By mapping the product categories to relevance on a per use case basis based on our engagements with customers, we were able to quantify the approximate allocation of the SDxN market sizing to each category of use cases. For more details on each of these use cases, please check out our use-case library pages on SDxCentral.com at https://www.sdxcentral.com/ sdn-nfv-use-cases/ For the 2015 report, the SDxN use cases are aggregated into the following categories and subcategories: • Network Access Control • Campus NAC • Branch NAC • M2M NAC • UC Optimization • Network Virtualization • DC Virtualized Networks • Campus/Branch Virtual Networks • DC Micro Segmentation • Network Function as a Service • Virtual Customer Edge • Virtual CPE (on-premise) • OTT vCPE (on-premise) • Virtual CE (telco) • Virtual CE (OTT) • Dynamic Interconnections • BWoD • Virtual Private Interconnect/Cloud Bursting • Dynamic Enterprise VPN • Cross-Domain Interconnect • Multi-layer Optimization • Virtual Core and Aggregation • vEPC, vIMS • vPE (inc vBNG & vCMTS) • GiLAN • Network Virtualization (Mobile) • Datacenter Optimization • Big Data Optimization • Flow optimization • Other (to capture other less common use cases not listed above) The use case model reflects fairly broad SDxN applicability. Of the 25 use case categories, 7 of them will contribute more than $5B annually by 2020, and the top 3 use case categories (what we term as umbrella use cases) are Network Virtualization at $28B, Virtual Core and Aggregation at $20B and Virtual Customer Edge at $18B.


Page 11

Market Report


SDX NETWORKING SPEND BY UMBRELLA USE CASES

$120


$100

$80

$60

$40

$20

$0 2015

2016

Network Access Control Virtual Core and Aggregation

2017 Network Virtualization

2018

2019

Virtual Customer Edge

Datacenter Optimization

2020 Dynamic Interconnects

Other

The Qualitative Impacts of SDN and NFV The quantitative impacts of NFV and SDN are substantial, but as we have observed over the past few years, the qualitative aspects of this market transition could be even more important. NFV and SDN marks the first time in the last 20 years that there has been a significant inflection point in both networking technologies and overall business landscape. Since we released the first report 2 years ago, we’ve seen and felt NFV’s and SDN’s impacts across the entire industry. We’ve seen new startups enter the networking space focused on both SDN and NFV and further acquisitions since the purchase of Nicira by VMware. Cisco, with its Insieme, Tail-F and other acquisitions, have rolled out various incarnations of its SDN strategy and has introduced (and declared end-of-life) controllers, some of which are based on open-source technology, while others other proprietary. Juniper, with Contrail, has proceeded down similar paths, mixing open-source strategy with proprietary products. All major networking vendors, including Ericsson, Brocade, Ciena, and HP have placed their SDN and NFV bets and it remains to be seen how the market will play out over the next few years. CLICK TO SHARE We’ve seen the threat of white box, gray box and brite box arrive on the scene, but THIS REPORT have yet to see massive replacements or significant margin erosions yet—though WITH OTHERS!


Page 12

Market Report

SDN and NFV Market Size Report 2015 many pundits believe the revolution is under way. Regardless, SDN and NFV have made their mark on networking, and the combination of new approaches, new technology and new players could disrupt what has been a virtual monopoly for the past 20 years. This will change how the major incumbents have to compete, and it could spur a massive consolidation of companies–both new entrants and established entities. Finally, given the number of systems that must integrate and the sheer complexity of the interactions between them, we will see a move towards higher-level orchestration and intent and policy-based systems. This, coupled with an increasing dependence on professional services and outsourced services will change even more how major IT vendors compete; opening the door for those who have strong systems integration practices. Regardless of the specific technology and strategic directions that the industry collectively pursues, the impacts of NFV and SDN are already significant and will continue to be over the next 20 years.

Conclusion The movement towards software-defined networks and network functions virtualization is the kind of change that happens once in a generation. But to think that the impact is primarily a technological shift would hugely underestimate the transformational impact that this will have on not only the networking space but also the entire IT infrastructure landscape. Just in the last two years, the number of end customers demanding SDN and NFV capabilities from their networking vendors has jumped dramatically as we had anticipated. And while massive SDN and NFV deployments are still not a common occurrence, reserved for the largest of organizations such as Facebook and Google, we are seeing early uptake across a wide variety of verticals, from finance to education. SDN and NFV have also ignited new rivalries and driven innovation at both incumbents and a new wave of startups trying to capitalize on this new level-playing field to disrupt two decades of incumbency. As the large players try to adjust to a swifter competitive environment, there will undoubtedly be a round of industry consolidation. Some companies will fail, others will break through, and still others will be acquired. The best interim measure of NFV and SDN progress will be the amount of revenue that is directly attributed to NFV and SDN as a primary selection criterion, and ultimately the number of NFV and SDN enabled networking nodes in production networks.

Contact Us If you have any questions about the model used in the report, or would like more details around the numbers, or are interested in a custom market sizing for your organization, please contact us at [email protected].


Page 13

SDNCentral, LLC 955 Benecia Avenue Sunnyvale, CA 94085 USA www.sdxcentral.com

The Trusted News and Resource Site for SDx, SDN, NFV, Cloud and Virtualization Infrastructure Rev A

SDxCentral SDN NFV Market Size Report 2015 A

Recommend Documents