On September 24, 2014, at the NASDAQ headquarters in Times Square, New York City, Brocade held its annual Investor Day conference. On hand were all the executives including CEO Lloyd Carney.
As expected, all the presentations presented a very upbeat characterization of Brocade. The major theme that Brocade wanted to set was the idea of “The New IP.”
The New IP
Among the major drivers of “The New IP” were the usual suspects identified by most IT vendors – big data, cloud, social and mobile. But one thing that CEO Lloyd Carney called out about what he calls “tectonic shifts” is that these shifts are all network-centric. The fact these shifts are network-centric may seem obvious, but few people understand this explicitly. Although obviously influenced by Brocade’s perspective, Carney’s his observation is nonetheless correct.
The shift from highly centralized (read: mainframe/mini) computing to distributed computing certainly highlighted the importance of the network. Sun’s Scott McNealy is famously and oft quoted as saying “The network is the computer.” But the reality was, that although networks were important, there was still a very dominant one-to-one relationship between the compute and the data.
However, in the 21st century, the big four pillars of technology – cloud, big data, social and mobile – do not rely on definitive or singular relationships between compute and data. In fact, the proliferation of data and available compute services (whether it is on-premise or in the cloud) means that the free-flow of data is not only desired, but necessary. So in order for smartphones using social networks to reach big data services over the cloud, each step requires data to travel over numerous networks. Each step is codependent on another step’s success. The steps of one user are interleaved with millions of other users to form a fabric or mesh of highly codependent steps in which the failure of one of the four pillars could result in interruption of services for the other three.
A good example is the impact a temporary internet outage has on even a single household. Imagine a “typical” modern-day family. It is likely that the parents rely on email for communicating between each other and friends, family and caregivers; the children may rely on the internet for homework research and gaming; and the family collectively may subscribe to one of the multitudes of video-on-demand services. An internet outage could disrupt all of these things! While not critical in nature, it can certainly be a nuisance.
This example placed in an enterprise context could be much more dramatic (and potentially disastrous to the business). An outage to the internet connection could result in the inability to communicate with customers, and adding fuel to the fire, it could disrupt the ability to transact business with customers.
The world today is truly network-centric.
The New Opportunity
One consistent message across all the presenters was the opportunity is net new to Brocade, and poses no threat of cannibalizing its existing business. This is true.
However, the new space into which Brocade is moving is already highly competitive. Cisco, Juniper and others have been playing in the space for many years. Carney and his other executives, many of whom came from Juniper, believe that their previous success and their relationships can be replicated at Brocade.
Results are mixed for those executives who move from one vendor to a competitive vendor and successfully convert customers to the new vendor.
However, the opportunity for Brocade is different. The opportunity for Brocade is new, well sort of. The new opportunity for Brocade is in IP-based SANs. This is all additive business for Brocade, which already dominates Fibre Channel (FC)-based SANs. While Cisco has been pushing FCoE, the adoption of FCoE has been anemic, and predominantly localized to bridging FC and IP networks.
Neuralytix believes and agrees with Brocade that there is a potentially tremendous opportunity for storage-specific IP-based SANs. These IP-based SANs will be driven by the needs of applications that move massive amounts of data around the internet. Some of it will be for processes such as ETL (extract-transform-load), a common big data process; others may include basic on-premise to cloud integration for replication, disaster recovery and/or business continuity; and accessing datasets foreign to an organization’s local datacenter.
These IP-based SANs, which for clarity sake in this research brief will be denoted by a different name – perhaps IP-based Data-Area-Networks (IPDANs). The reason is that, distinct from SANs which essentially carry blocks of data, these IPDANs will be moving enriched data objects. (Note, Neuralytix does not believe in the practice of introducing unnecessary new terms for marketing purposes only.)
Whatever they are called, Neuralytix’s initial research estimates that the growth in revenue and ports for IPDANs is in the strong double-digits. This outperforms spend on legacy IP spend by a whole order of magnitude (albeit from a very small base). Our initial research suggests that IPDANs are expected to make up a material part of all IP network spend by 2020.
Other characteristics of IPDANs is that traditional storage administrators are likely to “own” and manage these networks, in association with virtual server administrators, much like the FC-based SANs today. Traditional network administrators will continue to manage networks between the server and clients.
At the beginning of this section, we discussed that we believe Brocade will not cannibalize existing sales. We justify this on the basis that much of the existing FC-based SAN sales are deployed in OLTP and RDBMS environments. These environments, although they can be run on IP-based storage, are optimized for block storage over a more efficient, reliable protocol such as FC.
The market is still nascent; the opportunities still new. What is in Brocade’s favor is that taking an early mover advantage into a storage networking market is not new for them. Their leadership in FC is a result of their early-mover advantage in the mid-1990s.
The major challenge will be to see whether users, at a time when convergence is preferred, are willing to take a bifurcated approach to their IP networking.
The sales team at Brocade should dust off playbooks from 20 years ago, and learn from those who came before them. Although the world has change dramatically, Neuralytix believes that marketing “The New IP” will be just like marketing “Ye Olde FC.”