Hoov's Musings (volume 3, number 7)

What's Hot (And What's Not)
Mark Hoover, President, Acuitive, Inc.

There are a lot of interesting things going on in networking these days. By any criterion; the number of networking start-ups, the types of services being projected, the volume of evolving standards, the projections of future traffic volumes, etc., it is easy to identify networking as the “in” thing these days. It used to be iron, then oil/steel, then computers, and now it is networking.

But – if you are watching the industry, or even if you are involved in it on a daily basis, it can be difficult to separate the Big Trends from the small ones. We are bombarded with information on a daily basis that calibrates the day-to-day incremental changes that are occurring. But this information can be as insignificant as the reported movements of the slug under the bark on the tree in the forest, or it can be an indicator of a major trend that will change everything.

So what are the Big Trends? I’m pretty much convinced that the world of networking infrastructure, in terms of exciting areas to work in that will potentially bring a lot of value to the marketplace, and presumably riches to those who deliver that value, has settled into five key areas.

Optical networking
Mobile wireless packet services
Internet traffic management
New generation content servers
B2B infrastructure

This isn’t particularly insightful, I know. Each of these broad areas encompasses many tens of different product categories, hundreds of vendors, thousands of price list items, and tens of billions of dollars of market opportunity.

Optical networking ranges from core transmission systems to bandwidth-provisioning cross connects, to IP-aware ADMs, to edge and core lambda-aware routers, to SONET and Ethernet-based MAN solutions – and for purposes of brevity, I’m leaving a lot of stuff out. The area is rich in opportunity for both systems and component companies. The end game, in terms of commonly deployed architectures, is unknown right now. The only known is that the development and deployment of these solutions will change the way we think about bandwidth in the core of the network – it will become and is already becoming a readily available and cheap commodity resource. An even bigger unknown is how various service providers will incrementally migrate to the unknown architecture of the future. Therein lies the future reason for the success and failure of a whole lot of companies, whose prospects cannot easily be evaluated until a few years from now.

Lack of mobility is the greatest regulator to further expansion of the use of networked services. But the bar should be raised dramatically in the coming years. New technologies such as GPRS, EDGE, and 3G should enable an ever increasing amount of packet-oriented bandwidth connectivity from the optical/IP world to a wide variety of mobile handsets such as laptops, cellular phones, PDAs, internet radios, etc. I would bet that in five years there will be 5-10 times more wireless devices receiving and sending IP packets than there are wired PCs in the world in the same time frame. This implies a huge market for handheld devices, upgraded cellular systems, and infrastructure that connects the wired world to the wireless world, and applications. This is the one area, of all the Big Trends, where the big traditional Telco vendors are taking the lead. Excepting the applications side of things, there are only a few innovative and fast moving start-up companies participating in this space. This is based on the assumption that the buying consumers will be the existing wireless operators, and that they will buy only from the “safe” big guys. Personally, I think those assumptions are wrong and that there will be a lot of opportunity for clever and innovative start-ups that target this space. The ones that already are may be the some of the hottest names in networking soon.

The term Internet traffic management encompasses a broad range of functions and product types. Server load balancers, content switches, caches, bandwidth managers, firewalls, VPN devices, security (encryption) accelerators, intrusion detection devices, service creation and service level management, content management, etc. These devices can basically be put into two broad categories, those that worry about security and those that worry about regulating the flow of traffic. It is the latter category that spawns a wide variety of “edge” devices. These “edge” devices are generally oriented towards giving their owner the ability to determine service levels by managing the traffic flows between low speed peripheral networks or end systems and the high-speed core network. Where that confluence occurs defines the “edge.” And there are a lot of edges. Various vendors have postulated a couple of dozen or more, all validated by at least a couple of customers (usually service providers). I believe that the vendors that succeed and endure in this space will be the ones who have come up with the best definition of “edge.” It has to be a location where there is a serious and enduring resource (for example, bandwidth) mismatch. I italicize the enduring part because many locations where there is a paucity of bandwidth today, or a big mismatch today, will go away tomorrow as optical networking gets deployed more widely. Thus many vendors in this space will see their “edge” go away. The one area I don’t see this happening in is at the network/end system interface. Networks are getting to be much faster than servers, caches, and appliance-based networking devices. So Internet traffic management devices that focus on Virtual Resource Management, i.e. making multiple web servers, application servers, databases, firewalls, VPN devices, caches, etc. look to the outside world like one large virtual machine will thrive in the coming years. ASPs, CSDs, content providers, and dot-com’d enterprises will be the biggest users and benefactors of this technology.

That gets us into the related area of network-attached storage. For years, the dirty little secret of networking has been that end-to-end performance has little to do with the network, but mostly to do with the ability of the end systems to efficiently retrieve content from the storage subsystems. Delivering content means accessing information stored in optical jukeboxes, RAID disk arrays and other forms of storage media and getting it onto the network. Traditionally, servers optimized for accessing storage subsystems for internal computation purposes have been used as storage-to-network gateways (servers). That promulgates an inefficiency that needs to change. So a new category of device that will thrive will be new forms of servers, optimized to the functions of being a web page/object content server, or video server, or multimedia server, or XML server, accelerated by network processor and dedicated hardware assist for content/packet processing as bits travel from the storage-optimized technology domain to the network-optimized domain. Look for the entire server market to be turned on its head over the next 2-3 years

All this networking technology has to have a financial driver and the main one to do is B2B applications. Many financial analysts believe that the extended bull market we are in is in large part fueled by efficiencies that businesses have gained by leveraging computing and networking into their business processes. B2B is just the latest and perhaps most powerful wave of that trend. Which means that mind bending amounts of money – possibly trillions of dollars – are on the table, just from the cost savings associated with paper elimination, accelerated business interactions, and just-in-time supply chain management for all types of businesses – not just technology-oriented ones. I don’t know too much about exchange sites and B2B applications, but I do have some insight into the opportunity for those that build enabling infrastructure over which those applications run. So do about 500 vendors of various software products, all of whom claim to be (and justly, I believe) in the B2B space. Segmentation here is critical to the success of any vendor. And consolidation has started to occur. Areas that were distinctly separate product categories in 1999 are expected as integrated feature sets in 2000 and that trend will rapidly continue. And another trend will continue and accelerate – it’s all about XML. Get to know it and get to love it.

If you are in a business operating in any these spaces, you get the feeling right now that there is so much opportunity, and so much investment pouring into them, that you cannot help but succeed. In other words, all boats rise as the tide of investment flows in. That’s a dangerous mindset however, one still can succeed or fail based on having the proper market segmentation, problem definition, solution differentiation, and execution. But if you get it right, the rewards can be tremendous. At Acuitive we are oriented our projects and client profile more and more to line up with these key areas.

That’s not to say there aren’t opportunities in other areas, such as IP-based voice services, voice recognition, fixed broadband wireless, video delivery systems, metro-are network solutions, off-campus Ethernet switches, ASPs/MSPs/xSPs, last-mile broadband technologies, storage area networks, content management, network and systems management, campus networking solutions, core routers, and on and on and on. It’s just that in these areas you don’t have as many mega-trends working for you. You have to segment a little finer, work a little harder, and execute a little better in order to succeed. But there are plenty of examples of people doing just that. So, at Acuitive, we don’t reject these opportunities. We just examine them a little closer before we jump in.

(volume 3, number 7)

Home

Clients

Services

Hoov's Musings

Research Reports

About Acuitive

What's Hot (And What's Not) Mark Hoover, President, Acuitive, Inc.

What's Hot (And What's Not)
Mark Hoover, President, Acuitive, Inc.