Beyond Category 6: What's the big deal?
It may be that most customers perceive no compelling need for anything beyond Category 6.
Have you ever tried the pickup line: "Hey, I make copper cables for a living; what do you do?" I would not suggest it. And don't think substituting the word "fiber" for "copper" will get you any more respect—even that has lost its luster over the years.
To the outside observer, the world of wire and cable would seem as exciting as watching Prometheus endlessly push a rock up a hill. But after 15 years of doing nothing but connectors, wire, and cable, I can say with authority that the world of structured cabling is a fun and interesting place in which to be. What follows is an insider's history of the industry and some thoughts for the future. So, for those of you who actually find the world of cabling infrastructure interesting, I invite you to read on.
The many faces of UTP
Not long ago, it seemed every system had its own specific cabling and connector type. Through standardization, many of those variants have since become nothing more than conversation pieces lying unused in customer ceilings. We have seen a plethora (always wanted to use that word in an article) of beautifully designed, high-bandwidth cable iterations wind down to one dominant variant—unshielded twisted pair (UTP) cable. Whether it is Ethernet, Token Ring, or ATM (Asynchronous Transfer Mode), UTP has found its dominance within our marketplace.
But the single-variant concept of UTP has had a few bumps along the way. In fact, during the middle '90s, there were no fewer than 20 different varieties of UTP cabling available for install. Every electrical stricture from "A" (attenuation) to "Z" (impedance) was exploited. With endless parameters with which to differentiate, new products were hitting the market every few months or so. The pace was so rapid that marketing departments were full of people dedicated to finding ways of improving upon names that already had "+" signs or the word "super" next to them.
Of course, this product proliferation ran counter to the whole standardization process. Fortunately, there was a team of individuals who set out to make sense of it all by classing multiple cable varieties into single "levels" of cable. Some of these performance levels became the basis for further industry standards, which would eventually allow for the consistent transmission of data rates approaching—even surpassing—a gigabit per second. Ultimately, this would lead the industry back toward a few distinct cable variants, presently known as Category 5e and Category 6.
Both cable varieties currently act as a transport for some of the most popular protocols available today. Although Category 5e can handle the same top speeds as its Category 6 counterpart, it is the increased bandwidth potential of Category 6 that creates its strong market share.
The most important reason for this potential is what I call the 15-year issue. We can absolutely argue that the "type" of UTP cable being used in the prior 15 years has gone through at least one iteration (remember when Category 5e didn't have an "e" at the end of it?). Applying our history lesson tells us that Category 5e could also undergo an additional iteration some time over the next 15 years. Fortunately, the "next" iteration is already available—Category 6. So, people who are afraid they might get boxed in are opting for a Category 6 solution.
There is also a significant installed base of Category 6 already in place, increasing the motivation for hardware manufacturers to take advantage of its enhanced performance when compared to Category 5e cables. And finally, it is a standard—the benefits of which need no explanation. Most users who employ Category 6 are those who own or intend to occupy a building for an extended period of time, or are pushing the limits of bandwidth availability.
Beyond Category 6?
But now that the "next generation" is available and standardized, the question begs to be asked: "Is there a need to go beyond Category 6?" It appears there are many manufacturers in the industry who think so. Recently, I have seen a class of cables people have labeled "Category 6e." (Please note that there is no Category 6e standard; I am simply referring to a term being used by the industry as a way to identify cables that are more advanced than Category 6.) At last count, there were no fewer than six different manufacturers touting this capability. Between the time this article is written and the time it reaches your desk, that number may very well grow.
But claiming availability is one thing; determining customer need is entirely another. Not too long ago, I was talking to a customer about the very topic of Category 6e. The customer asked if we were "supporting" this Category 6e classification. I was straightforward in explaining to him that we did not yet see the benefits of offering such a solution. My great insight came when he reacted with relief instead of disappointment. He said one of our competitors was trying to get his attention by touting a new "great" solution, namely Category 6e. They talked to him about how improved and advanced their cable was compared to current industry standards. After listening intently, the customer explained to us that he asked the manufacturer a very simple question—one they could not adequately answer: "So what?" To paraphrase, "What can a Category 6e cable do that a Category 6 cable, or for that matter, a Category 5e cable, cannot?"
I instantly understood his point. When all is said and done, will the extra headroom obtained by using a Category 6e product improve the customer's current or foreseeable network performance? The manufacturer certainly had no proof that it would. And with that, the customer's conclusion was clear. The extra headroom over Category 6 would make little or no difference in their system performance.
The customer brought up other points, including lack of support for a Category 6e cable from the active equipment manufacturers, and the nonexistence of any standard committee draft stating what an unshielded Category 6e electrical specification sheet would look like. The only thing the customer seemed sure of was that Category 6e cables would be more expensive.
Brave new world
Does this mean that we will never see a Category 6e cable? I am not saying that at all, but it is clear to me that most customers are not yet seeing the value this type of solution provides. Let me explain. For the past 15 years or so, we have been able to turn a simple 4-pair telephone cable into a high-end, high-speed data-throughput "machine." With each new innovation, eager customers lined up, willing to buy into the latest technology.
But we are now in a different world. LAN topologies have become predictable, and the products they use are standardized and available. In other words, people know what they want (and need). This is especially true for the wire and cable portion of the LAN. Most customers understand that they "need" Category 5e, and for those who feel a little uneasy about it lasting 15 years in the ceiling, they have a Category 6 alternative. Most customers perceive no compelling need for anything beyond Category 6.
Another factor slowing demand for infrastructure other than Category 5e or Category 6 is the current state of the economy. Companies are now in a mode of challenging their IT departments to delay or even cancel upcoming infrastructure improvements. Any upgrade to new or premium cabling solutions must be justified by immediate "today" benefits. This, coupled with the trend of centralizing employees by closing and consolidating locations, supports the opinion that existing infrastructure will be supplemented rather than changed.
Although these points tell us a true Category 6e might be a ways off, facts also have implications for trends in near-term product innovations. I would look for a whole new group of differentiated products that will give customers immediate benefits. It is clear to me that customers are now, more than ever, demanding product advantages they can take to the bank now. These attributes include product size and cost, implementation cost, data-throughput advantages, and procurement advantages.
Another strong differentiation will come in the form of "safer" cables. Whether it is cables that will "take the lead out" or cables fabricated with compounds that burn and smoke significantly less, this will be of great interest in the coming months and years. All of these innovations can occur within the overall performance levels of Category 5e and Category 6.
You might have noticed that I have not talked much about two important infrastructure variants: fiber and wireless. Since my conversation has been mostly geared toward desktop applications, I have purposely steered away from these options, for two reasons.
First, there is little motivation to invest in replacing an existing cabling infrastructure type with a totally new one. Let's face it, most end users are running 10/100 Mbits/sec to the desktop and preparing for an eventual switch to gigabit. Fortunately, the vast majority of the installed copper base will support this speed increase. This fact alone would suggest that a "forklift upgrade" (total replacement of existing copper infrastructure to a fiber infrastructure, for instance) would be difficult to justify.
Second is cost. Fiber makes a wonderful backbone. Wireless does great between buildings. These are prime examples of places where dozens, if not hundreds, of signals are traversing through essentially one conduit, and the need for high bandwidth is readily evident. People soon learned that a few fiber strands could accomplish as much as a 25-pair copper backbone cable. This meant that the cost of the fiber (or wireless) system was offset by reducing infrastructure needed to run simultaneous signals.
Accordingly, fiber-in-the-backbone is the rule, rather than the exception. But with desktop applications, there is only one end user using limited data streams. When fiber goes against copper where high bandwidth-carrying multiple signals is not necessary, cost typically works in favor of copper.
Now this doesn't mean there won't be some portions of specialized networks that use fiber and wireless to the desk. The flexibility of wireless and the high-bandwidth capabilities of fiber make them an important part of LAN installations—especially in vertical applications, such as schools and universities. But with that said, both applications with respect to desktop use remain a niche for most people—a very important niche, but not the total solution that will be running to every cubicle in the next one to three years.
A little further outU
Here is where it gets fun, and where putting my thoughts on paper may get me in trouble.
Our industry faces some interesting questions as we look out three to ten years from now. If trends continue on their current path, Gigabit Ethernet-to-the-desktop will become commonplace within this time period. Inevitably, people will then begin to think about 10-Gigabit systems to the desktop. The question is, will copper be able to handle 10-Gigabit practically? I stress the term "practically," because for 10-Gigabit to work over copper, it also needs to be cost-effective and backward compatible. And by the time customers really start thinking about 10-Gigabit, this problem must be solved or copper's days are limited.
That said, there is ample motivation to solve it. There are 20 billion feet of Category 5e and Category 6 installed-base infrastructure out there. For active-equipment manufacturers to come out with equipment that ignores that installed base means ignoring their biggest potential market. This single point is why I think 10-Gigabit eventually will work over some variant of copper. The "home run" would obviously be the ability to run over both categories of copper (5e and 6).
If, on the other hand, 10-Gigabit can't be made to work on copper, I think we will see the same mix of direct line versus wireless—but the direct line will be fiber. Of course, it will take several years, if not a decade, for that transition to happen. But if copper can't do 10-Gigabit at a reasonable cost, then it's only a matter of time. It is possible we will see an approach that involves incremental bandwidth strategy between 1 and 10 Gigabit over copper, but eventually, the 10-Gigabit question will need to be answered.
And what about wireless? I hear a lot about security, line-of-sight, and other issues. But in my opinion, those are all solvable issues. In fact, people who use wireless tend to swear by it. The flexibility is quite beneficial. The real factor, again in my opinion, is cost-effective bandwidth. Wireless needs two things to be a serious contender to replace copper in the next three to ten years: Bandwidth up to 1 gigabit (with the potential for more), coupled with a desktop solution at a cost comparable to what people can buy fiber/copper for today.
Failing that, while wireless will always have a place for those who want mobile access (such as the education vertical market), it will face difficult barriers for high-bandwidth applications.
The road ahead
It is fairly certain that our industry will enjoy several years of product-performance stability. Category 5e and Category 6 will remain the cable of choice. For an industry that has focused primarily on the next "number" (Category 3/4/5/6) for the last 15 years, having some stability is not such a bad thing. That doesn't mean innovation is dead. Let me reiterate the near-term explosion of products discussed earlier that are directed at satisfying "current" needs, such as safety, cost, and ease of installation, providing benefits realized in months rather than years. These innovations will happen within the confines of basic electrical standards. It is certainly the direction our company is taking seriously.
And regarding the future? As previously stated, it all will focus on which infrastructure—copper, fiber, wireless—will run 10-Gigabit most effectively at the best overall cost to the end user. When that question is answered, the end-user choice for infrastructure will be clearer than ever before.
Bob Kenny is vice president of Krone Inc. (www.kroneamericas.com).