In 2015, The Rayno Report teamed up with SDxCentral to bring you the latest news, trends, and analysis of software-defined everything (SDx) markets. As I came up to speed on everything software-defined, it was an adventure to learn about new markets, new companies, and new people.
I look at writing year-in-review pieces as an exercise in discovery — by reading dozens of the year’s articles and going through my notes, I often stumble upon things I had forgotten about or that somehow can be viewed in a different perspective with the passing of time. So, I’ve reviewed the year and now I’ve got my list of things I learned about SDx markets in 2015.
A recurring theme is that big changes are at hand in both networking and data center markets as the world has moved to a cloud architecture, where data and applications can be shifted around on a moment’s notice, making the physical home of the service meaningless. This is, of course, the concept known as virtualization, whereby applications are detached or disaggregated from the underlying hardware. The most important aspect of this is on the development level, because it has ushered in the era of agile development in which software can be designed, deployed, moved, and updated on the fly.
Here’s what I learned about what’s going on in 2015:
1. The Stakes Are High in SDx
The first bump in SDx came with the emergence of the software-defined networking (SDN) market. The SDN hype was started by the introduction of the OpenFlow protocol, which was first released on December 31, 2009. A number of startups started down the path of SDN, including the mighty Nicira — still the first and only billion-dollar exit in the SDN startup world.
After seeing the success of Nicira, Cisco realized it had to get in on the action, and in April 2012 Cisco announced it had invested $100 million in Insieme, which was structured as a spin-in with Cisco getting the option to buy. Later that year, in July, VMware announced it was paying $1.2 billion to acquire Nicira. The battle of the network virtualization giants was born. Cisco closed the Insieme loop by purchasing the company for $860 million in 2013 to form the basis of its Applications Centric Infrastructure (ACI).
But the SDx market has grown well beyond that early vision of data-center virtualization. It now extends out to the WAN, with SD-WAN and virtual CPE. It includes cloud security, whether it be network microsegmentation, analytics, or even container-based security technology. And there’s an entire market for service providers known as network functions virtualization (NFV), which takes a purpose-built communications infrastructure and converts it into software.
When you add up the market sizes, the total addressable market for SDx is $130 billion by 2020, according to our research. Telcos are likely to spend upwards of $25 billion on NFV and SDN technologies. Of course, some of this is “replacement” dollars, in which legacy hardware technology is being replaced by software. But some of this spending is new, as well.
That’s why there is a lot of money chasing these markets. Our research indicates that at least $4 billion has been invested in SDx startups. That breaks down as $700 million in SD-WAN, more than $600 million in cloud security, and at least $2 billion in network virtualization and data-center virtualization.
All this amounts to huge stakes — for the customers, the investors, and the suppliers.
2. There’s a Cloud Security Bubble — But We Need It
As I mentioned, more than $600 million has been invested in security startups. Some of these companies have become “unicorns” — startups with more than $1 billion valuations. There’s also a host of recent cloud security IPOs whose share prices are zooming in the aftermarket.
What’s it all mean? The ongoing cloud security bubble isn’t irrational — after all, when you read about major security breaches every day, it’s clear there is huge demand for security software. With the boardrooms of Fortune 500 companies panicking over data and application security, there’s no reason for it to slow down any time soon.
Bubbles have value in themselves even though there is going to be pain and carnage along the way. Even as a herd of startups is culled, the bubble can accelerate innovation in specific markets. What’s interesting about the orginal Internet bubble, which resulted in a crash and many failed companies, is that it created the largest economic engine in the world — and some of the world’s most valuable companies, including Amazon and Google. The same thing is happening in SDx and cloud-based security. The bubble will drive more innovation in the market.
3. NFV Became a Big Deal
A funny thing happened in the SDx market. The global service providers — those plodding behemoths that deliver our communications networks — got turned on to SDN and decided they needed their own carrier-grade infrastructure to implement SDx concepts.
In 2011, the European Telecommunications Standards Institute (ETSI) developed the management and orchestration (MANO) model to define and develop standards for NFV in service provider networks. This MANO model has taken on a life of its own, and continues to be referenced as the primary carrier NFV model, although other groups such as OPNFV and vendor-specific software extensions of OpenStack are looking to fill in gaps.
Whether you’re a MANO fan or not, the service providers have made it clear that they are experimenting with NFV and are encouraged by the concept — though there are still many critics who say the business cases for NFV have yet to be established.
In 2015, we learned that the service providers were willing to stand up and speak about NFV and its possible applications. 2016 will be a crucial year for NFV, when the operators start deploying the technology and finding out whether it proves out the ROI and business case or not.
4. John Chambers Picked the Perfect Moment to Retire
It was fun to get involved in the John Chambers retirement prediction sweepstakes. After years of deliberation, he finally pulled the trigger. And it seems as if Chambers, the CEO at Cisco for two decades, picked the right time.
First of all: 20 years is a nice tenure. Chambers built Cisco into the networking powerhouse it is today by acquiring a series of Ethernet switching companies and then using that base to extend into adjacent markets such as data center, security, and services. Chambers also guided Cisco’s SDx strategy with ACI and the Insieme aquisition, and decided that once that strategic direction had been set, it was time to kick himself up to the executive chairman’s office. Cisco remains one of the world’s most profitable technology companies. And while its recent share-price gains haven’t been anything to write home about, Chambers managed to boost the stock price and attract a new class of value investor by upping the dividend share buybacks.
In the end, the timing was right for Chambers’s retirement as CEO. He leaves a solid legacy as one of the most reliable technology CEOs. Now it’s on new CEO Chuck Robbins to take Cisco to the next level, which will be challenging, given the law of large numbers and the fact that the network and service provider markets both have their inherent challenges.
5. Large Carriers Are Still Confused About What They Are
Large telecommunications service providers can provide great services, and they can also be maddeningly bureaucratic organizations, like governments. Customer service in the service provider market is among the lowest of all consumer industries: Telecommunications and cable operations consistently rank lowest in customer satisfaction.
The business trends are more problematic. First, there’s the fact that the service provider debt binge continues — especially in North America. As I pointed out, AT&T and Verizon alone have collectively increased their debt load by $100 billion in the last year.
And for what? Much of it has gone to spending on a patchwork of assets that may or may not help their profitability. The service providers have also indicated they want to head in the direction of becoming digital media companies. AT&T bought DirecTV and Verizon bought AOL. Both Verizon and AT&T seem to have been infatuated with content lately, perhaps seeing the success of their cable competitors.
Can telcos possibly change their culture to be customer-focused consumer media companies? I don’t think so. Despite their ambitions to become content companies, more innovative companies such as Netflix continue to beat them in the quest for the consumer’s heart. Service providers still haven’t been able to shed their “dumb pipe” image, and they don’t appear ready to do so.
6. It’s Possible to Ship a Live Giraffe Using SDN
In one of the most important discoveries of the year, I found out that it’s possible to ship a live giraffe through Canada using the Internet, and that it somehow all connects back to SDx.
Thanks to Netcracker for pointing this out at their analyst meeting.
7. Optical Will Be the Next SDN Frontier
Much of the SDN and NFV attention has been focused on the applications in the data center and the WAN, using switches and routers (virtual or not). But what about the core of the service provider network? This is where the bulk of networking traffic ends up, when your Netflix movies end up being switched as wavelengths on an optical box.
Managing bits on wavelengths can be driven by SDN technology. We’ve already seen some melding of SDN controllers with optical boxes. From an engineering perspective, service provider networks continue to be a messy combination of electronic switching technology (electrons through routers and switches) and optical switching technology (wavelengths). Over time the trend will be toward running more traffic through an optical-only domain. SDN promises to add intelligence, automation, and standardization to the optical networking market.
Many leading optical vendors are betting on this. 2015 saw a number of deals blending SDN and optical, including Ciena’s purchase of Cyan and the creation of the Blue Planet software division. Other optical vendors, including Infinera, have been talking about implementing SDN to automate the provisioning of packet-optical circuits.
Packet-optical technology, in which IP-based traffic can be packaged as wavelengths running over optical fiber, rather than electrical connections, offers much promise for SDN, because of its flexibility and efficiency. Wavelengths can be easily managed with software. The future of the optical network is one that can be instantly provisioned and automated using SDN control.
8. Megadeals Are Uninspiring
I’ve already talked about one megadeal, AT&T buying DirecTV. Many more have been in the works and more will happen — whether it’s the Cisco-Ericsson mega-partnership, Alcatel-Lucent merging with Nokia, or the next telco-cable deal. The service provider and networking vendor communities have been in an era of mega-consolidation for decades.
Both from a consumer and analyst perspective, this trend is uninspiring. It leads to less choice for the consumer and slower technology development in the industry. The fact is that large, multibillion and multinational service providers do not move that quickly.
But it’s not going to change. We’re going to see more of this in 2016 (Verizon buying Yahoo is the speculation du jour). These huge companies continue to push the limits of merging in order to satisfy shareholder needs. I must confess: It’s painful to watch, because in the end, consolidation means fewer jobs and usually slower innovation.
Those are some of my biggest takeaways from 2015. In the first few weeks of January, I’ll be writing more about the trends we expect to see unfold in 2016. For now, I’d like to wish all our readers Happy New Year!