Considering how long it usually takes the IEEE to ratify critical networking standards, its rapid standardization of the NBASE-T Alliance's development of 2.5 and 5 Gigabit Ethernet over Cat 5e and 6 copper cabling was surprising. In just under two years since the NBASE-T Alliance formed, the IEEE approved the alliance's recommendations and published the 802.3bz standard. Considering all the effort to accelerate industry standards for multi-gigabit Ethernet over copper cables, the demand for such a technology must be sky-high, right?
Well, the answer really depends on your specific situation. Many are quick to point out that 10 Gigabit Ethernet over copper will likely be the prevailing Ethernet technology as soon as hardware pricing comes down. Yet price drops are not the biggest obstacle for 10GBASE-T. Instead, the real problem resides with the type of copper cabling commonly deployed in enterprise networks today.
The majority of installed cabling is either Cat 5e or Cat 6. Up to 100 meters, both cabling standards can handle 1GbE. However, making a leap from 1 gigabit to 10 gigabit speeds with the same distance rating is impossible. The minimum cable type recommended for 10GBASE-T is Category 6A. Given the fact that Category 6A cabling must be able to operate at 500 MHz, those cables are rated twice as fast as Category 6 and five times as fast as Cat 5e cabling. Just looking at the difference in size and twisting between Category 6 and 6A cabling, you realize that the cables are not simply a minor revision. Instead, Cat 6A is completely different animal.
Considering that the cost to pull and certify thousands of new cables can be incredibly expensive, many enterprises are likely stuck with their current cable for the foreseeable future. The time when an organization finally budgets for a new cable install is likely years off. This is exactly why the industry is in such a hurry to make multi-gigabit Ethernet available to customers.
So, now that we’ve established why network vendors think that 2.5 and 5 Gbps Ethernet is going to be popular, let’s look at some reasons why organizations may want to consider multi-gigabit Ethernet.
The most popular rational for implementing multi-gigabit Ethernet over copper has to do with 802.11ac Wave 2 WiFi. With Wave 2 technology, you can theoretically reach maximum throughput on a wireless access point that goes beyond 1 Gbps. That means that the 1Gbps copper link that connects the AP to the LAN becomes a bottleneck. But if you replace your 1Gbps Ethernet switches with a multi-gigabit capable switch, you can now connect each AP at speeds of 2.5 or 5 Gbps depending on whether you have Cat 5e or Cat 6 cabling. While this indeed eliminates the theoretical bottleneck, some WiFi experts say Wave 2 doesn't require multi-gigabit backhaul links. There’s no doubt that 1Gbs and greater WiFi throughput on a per-AP basis will happen eventually, but it’s unlikely in the near term.
The other use case for multi-gigabit Ethernet is when your switching uplinks connect using copper cabling and those uplinks are becoming overutilized. Sure, you could perform link aggregation if multiple pulls are available between two switches. But if that’s not possible, you have the option of upgrading your switches to take advantage of multi-gigabit Ethernet over existing copper pulls.
The last and most logical rational for moving to multi-gigabit switches is a simple case of future-proofing. If you are about to perform a network switch refresh, you may want to look at NBASE-T and MGBASE-T-capable switch options. You’ll likely find that the additional cost of implementing these technologies isn’t tremendous when you consider a four- to five-year hardware lifecycle. Doing so helps to protect future WiFi technology deployments as well as any other bandwidth-intensive projects that might crop up. Since many enterprises are far away from any major re-cabling projects, it’s likely a smart investment.