(Note: It looks like the article has since been taken down 10/18/2014)
A recent and rather ambitious outlook on Cisco at Motley Fool entitled “Cisco and the Golden Age of the Internet “ talks about the rise in internet traffic and the growth potential it holds for companies like Cisco. While the article does mention competitive pressures (from the likes of IBM, HP, DELL and JNPR), it underestimates the impact of disruptive technologies on the burgeoning networking equipment space.
The network has finally become recognized for its importance in the post-PC era, as analysts and executives have come to grips with the impact of the hyper growth in IP addresses, streaming content, and the growing penetration of internet connectivity into broader arrays of once standalone devices and appliances.
Networks are larger and more complex today than they were perhaps ever intended to be, when Dan Lynch and other networking pioneers were creating early protocols. As discussed at Archimedius, this tension between initial network capability and ever-increasing demands has fueled the fortunes of many data networking companies who have transformed these emerging demands into ongoing product, support and channel revenue, as well as ongoing innovation.
In The Cloud and the Great Data Center Race I talked about the role of the network as one of three key areas of interest to CIOs migrating to private or hybrid cloud models. I pointed out that the preoccupation with hardware in the networking industry was slowing innovation, which could explain in part why Google has chosen to build its own network (“Going with the Flow”- Wired).
Google’s Primary Threat: As a Poster Child
But the hardware-centric networking industry isn’t threatened as much by Google as it is by the implications of what Google has accomplished and a vast new competitive landscape formed at least in part by the spread of virtualization into the data center. VMware, Citrix and Microsoft, as well as the public cloud players, have already disrupted the once powerful alliances between companies like Cisco and HP by changing the dominant ASIC hardware paradigm (once thought critical to data center infrastructure) into a more fluid service and software-centric model. HP can get into networking and Cisco can get into servers; something seen as impractical as early as seven years ago.
This is not a subtle issue but rather a strategic development that potentially undermines the ASIC-enabled moat that has protected the margins of the networking giants and their partners.
While the Motley Fool “Golden Age” piece does pay lip service to Cisco’s success with UCS and the “death of the PC”, it misses the significance of a data center transformation already underway that could erode a series of (perhaps once) competitive advantages for some players and make otherwise bit players ever more strategic to the evolution of the network..
For a perspective read The Big Switch to OpenFlow and Software is eating the world. Yes, there are a host of startups in the emerging x86 or software-defined networking space that may or may not survive. It remains to be seen whether or not they will join the graveyards of many other startups who have taken on Cisco directly.
Yet in 2012-2015 these startups stand a significantly greater chance of success than in earlier “speed and feed”, security and “VoIP emergence” eras- where they often would have had to make heavy silicon commitments ahead of buying cycles, and then had to compete with behemoths making one-off competition-driven discounts, thereby squeezing them out at the altar after they had also invested in sales and marketing.
Today the advantages of software may be harder to discount. Witness the quick success of Arista Networks, cited recently as a potential IPO as early as 2013.
Besides the nimbler world of software-centric networking (and how dynamic markets tend to favor more innovative startups), the core hardware-centric network is now competing directly with server-centric and services-centric players (including IBM and HP). This is more than a significant development, it is a harbinger.
Emerging Strategic Players
There is an unprecedented assortment of disruptive threats that puts even more emphasis on a host of players with in-depth knowledge, experience and software for various layers of the OSI stack (the core layers of technologies and protocols that allow networks to interact), including FFIV, Infoblox and RVBD. All are now strategic to the evolution of the network and the potential for software or hardware-centric automation and the further spread of virtualization as well as the adoption of private cloud operating models. Yet the rise of software has similar implications for all networking players, including these more specialized players.
For example, even network switch player Arista has the potential to play in “upstack” OSI areas like layers 4-7, the bedrock of FFIVs business. See /Arista-FPGA-switch-could-revolutionize-Layer-4-7-network-services. Yes, this new software-centric era could do the same thing for the various OSI experts as virtualization did for once partners Cisco and HP. Competitors become allies and allies become competitors.
Emerging startups as well as the various seasoned OSI layer experts just mentioned are now strategic to how the network evolves. On top of that companies like VMware, Dell, HP have the potential to disrupt the traditional hardware model that has served the networking leaders (and their channels) well, by cannibalizing the hardware-specific models (like what we saw years ago in the server industry) and replacing them with their own software running on commodity servers which are refreshed on a predictable basis. They could partner or acquire their way with these new players in order to breach the once formidable ASIC moat and offer new levels of scalability, agility and automation.
It is likely that we will see a new set of seasoned, well-funded and well-staffed startups with bigger threat potentials than before.
That explains the rumored launch of Insiemi (reported in the press in recent weeks) and the new rush for networking talent, including possible poaching from the new wave of startups. I’ve mentioned Arista Networks and Nicira, as well as OpenFlow before at Archimedius, but the reports of Cisco responding to the SDN threat (see Cisco’s mysterious spin-in hiring talent from SDN vendors – The Network Hub) seem to give this software disruption theory even more credence.
Another company of interest is NetYce, a company with a network management product spun-out of a European bank that is attracting new customers in Europe wanting to automate critical network management processes. I have only spoken briefly with them, but what they’ve said so far puts them in the interesting to watch category.
Whether these early stage companies survive as standalones or become a part of larger enterprises looking for an attractive entry point into the networking market, the race appears to be on.
The question for the established players like Cisco and Juniper is whether the growth in networking becomes a golden age or merely a golden fleece.
It is also likely that we will see an entirely new generation of IT infrastructure companies with more product/solution/service breadth than ever before, enabled by the power of software and perhaps a newer, more nimble combination of operating models. See this recent interview with Accenture’s Tim Jellison regarding the need for IT companies to have a multitude of operating models.
The biggest benefactor of the golden age may therefore be enterprises, and the tech companies nimble enough to meet their needs.
In May I’ll be moderating a panel of cloud experts at Future In Review 2012 on “What Every CIO Should Know About Cloud Computing.”