Data Centers

Will copper always be the networking standard?

Copper cabling has been the venerable media for IT for a long time. In this blog post, IT pro Rick Vanover shows reasons why copper will remain as the standard for a while.

There have been two popular posts on this blog about copper cabling that got me thinking about the future of this medium for networking. The first was a good refresher from Jack Wallen on how to create your own CAT-5 cables, and the other was Michael Kassner’s detailed information on how twisted pair cabling works. All of this is really good stuff, but I posed a question of why are we still using copper networking effectively the same as we have since the early 1990s?

Since that time, copper media has been the standard for inside the data center for server connectivity as well as to endpoint devices, such as PCs and printers. There are many reasons for copper media being successful, including:

  1. Low cost and interchangeable use: Copper media is very affordable for PCs, servers, and printers. Aside from any color preferences, it is versatile enough to work for all of these devices.
  2. Easy to make: Because copper networking is relatively easy to work with (compared to fiber), IT groups can make their own cables for specific length requirements, if needed, as well as reducing costs.
  3. Features keeping up: Twisted pair cabling has increased features over the years, most notably support for 10 Gigabit Ethernet and Power over Ethernet.
  4. Security concerns: While wireless technologies are increasing, most workplaces still don’t embrace them entirely at the endpoint and hardly at all in the data center.
  5. Distance: Copper media delivers a “good enough” distance for most situations; this may be simply because this is almost exclusively what we have worked with.
  6. Mixed compatibility: Revisions going back to category-3 cabling allowing backwards support for token ring and 10 megabit Ethernet networks are handled easily with backwards compatibility during periods of transition.

Even with these benefits, copper media still may not be the only cabling used. Fibre still reigns supreme in most large storage networks and wireless is gaining in the endpoint, primarily with new consumer-focused devices morphing into mainstream environments.

What will it take to move copper media from its stronghold? A new media doesn’t seem likely and wireless probably won’t scale and deliver the speed needed in the data center. I don’t see copper ever losing the throne. Share your comments on where you see copper media going.

About

Rick Vanover is a software strategy specialist for Veeam Software, based in Columbus, Ohio. Rick has years of IT experience and focuses on virtualization, Windows-based server administration, and system hardware.

66 comments
blackepyon01
blackepyon01

-Fiber optics for distances between large switches over 300'. -copper Cat5e for your network's backbone -wireless only for laptops and portable devices. -no wireless repeaters (they're bad news). Use wires to access points. At the school I work at, we have 30 laptops for student usage. All student computers are domained, which is painfully slow and problematic over wireless. Experience will say: wired is always better.

Oz_Media
Oz_Media

Fibre is obviously the current fast standard, I've implemented fibre in two locations, where it meant digging up the downtown streets to drop loops but the cost is astronomical. Street closures, city planning, construction, road rebuilding etc. It's an operational and statistical nightmare. During deregulation of LD in Canda, Sprint Canada (formerly CallNet Enterprises, not Sprint US) was installing a fibre based Sonnet loop network between Alberta and BC, meawhile I noticed that Ted Rogers was digging up a small community to install copper due to the lower cost. At that time Ted Rogers still operated a rapidly failing Unitel (before the Rogers Communications we know today). Even after this expansion, Unitel continued to lose tens of millions each day, he simply couldn't afford to keep up due to the cost of fibre and his outdated network. So clearly the future (back then) was seen in BURIED fibre. Unitel couldn't afford to compete anymore due to operating costs and was working an old dialer system as opposed to joining in the fibre carriers to gain equal ease of access (direct dialing). Companies that install fibre now have immense budgets, it's not for everyone. Not only the initial cost of fibre but the fragility of it, makes it a rare breed for internal 'exposed' applications. If you have enough money to integrate fibre, it is usually buried to protect investment, which generally excludes it from being practical for internal office use beyond the demarc point. I think that until a more coest effective and DURABLE solution is found, copper will always be the most commonly used.

AnsuGisalas
AnsuGisalas

Twisted pair wire get accidentally subjected to a very tight bend? No problem. Less so with other cables, or am I behind the times?

Kevin@Quealy.net
Kevin@Quealy.net

... but I wouldn't be surprised to see it stick around for another 50 or even 100 years until another alternative comes along.

techbb
techbb

I have recently been experimenting with ethernet over mains. I have concerns about security but does this technology present a realistic alternative?

CharlieSpencer
CharlieSpencer

Faster wireless or cheaper fiber, take your pick.

Slayer_
Slayer_

We need that stuff to be invented. That would probably give us a huge leap forward in technology.

Papa_Bill
Papa_Bill

Even being bent only once, copper loses some of its density due to fracturing of the lattice. This will increase resistance at that point and contribute noise. Not much, but some. There are limitations to the bend radius in the installation standards (as well as for fiber), I don't recall just what that limit is.

Oz_Media
Oz_Media

Fibre, besides being extremely expensive to drop, is also extremely fragile and has limited applications. That's why it is generally run beside railroad tracks and buried. Not only is it easier to drop from a train (especially in Canada where crown corporations once monopolized the railroads and phone carriers), it is in a heavily protected environment where the ground is more stable. In the city it is buried besides sewer lines for added structural integrity, but the alley and grassy back yard? Forget it, who has the money to invest in running fibre to residential areas? All we see is a loop dropped into the business centres of major cities which can cost hundreds of millions if not billions depending on the size and logistics. Carriers are not about to drop a line in your residential neighbourhood any time soon.

b4real
b4real

I wonder if a new blend of metals also can get the same distance with lesser expensive metals.

harryolden
harryolden

All things change but not before that the backyard job can be done by the handy man to lay fibro cables in his home untill then it is too expensive, but dont worry sombody will work it out on how to join the cables on the cheap.

seanferd
seanferd

You can buy a fairly cheap bit of hardware what plugs into a wall socket. can't recall any of the names. I wouldn't use it in an enterprise environment.

johnm
johnm

In theory (from a long time ago) you should be able to get relatively low-speed data transfers over mains for a hundred feet or more. The data will transfer along your mains wire until it reaches a transformer. The transformer will not pass the higher-frequency data signals to the source side. So, all of your data-sharing devices need to be getting their power from the same trandformer. Conversely, anybody else drawing from the same mains transformer will have access to your signals and can read them with the proper equipment - or they may inadvertently put their own data on the same mains as you while using some or all of the same frequencies and interfere with your data transmissions. If you are the only one on your transformer, it shouldn't be a problem, but you may not have control over that. I haven't looked at that tech for decades, but I'm sure that they have ways now to use multiple frequencies to increase data rate - we were working with words-per-minute, not megabytes-per-second.

b4real
b4real

Power? Or Water?

Papa_Bill
Papa_Bill

without the use of metallic (i.e. copper) conductors?

Papa_Bill
Papa_Bill

Due to the ever expanding traffic load, frequency and bandwidth limitation and increasing interference.

bboyd
bboyd

more flexible fiber, wireless with less overhead (bandwidth losses to format)

pgit
pgit

You are correct so far as price-point on the fiber. Still, I won't be installing any anytime soon, nor throwing cable away and setting up wireless hubs. Not just speed, but reliability is a huge issue. I had a client with a critical networked app that simply would not take wireless. It has a central DB that's pretty time-intolerant so far as latency. We can't even have multiple switches in the mix. Fiber would help here immensely. But guess what? It works now. There is no justification to spend $$ on what ain't broken. Most other clients of mine don't want wireless for plain security reasons. They don't even want people to know they have a computer network running in the place, ideally. Broadcasting the fact, and potentially having data jacked is a non-starter. Out here in depressionland I don't see anyone getting excited for any of the recent developments, USB 3, Light Peak etc. The price of fiber could drop to free plus installation and none of my clients would be interested at the moment.

seanferd
seanferd

copper goes way up again. Further along, as the very architectures of computing change, fiber may become more sensible than copper, depending on which way the chips fall. There are always transducers, so it may not matter at all, but there may be payoffs going one way or the other depending on whether spintronics- or optical-based architecture takes hold. (Only two options out of potentially many.) But you pretty much have it nailed, I think, for the near future.

HAL 9000
HAL 9000

Aluminum is a extremely energy intensive material to make and it's always going to be expensive no matter how much it gets subsidized. There is a reason why so little of it was used in the early years where cost of Production was sort of important. ;) Col

Oz_Media
Oz_Media

A set standard will ignore key issues such as type of copper, insulation used, manufacturing process etc., which all cpntributes to conductivity and flexibility also. Copper wire isn't just copper wire.

AnsuGisalas
AnsuGisalas

as I recall though, fiber goes permanently opaque, or breaks outright, if it is bent too tight. Copper is pretty sturdy. Of course, for bendability you'd have to go for wireless, but that's a different problem...

Papa_Bill
Papa_Bill

from pole-to-pole with a low-cost adaptor coupling it to the customer's drop, and our love-buddy AT&T gives us 'phone. Internet and 300 channel DTV on a single drop from fiber on the pole.

HAL 9000
HAL 9000

Here we'll have Fiber to the Door in the next 8 years to over 93% of the population. Depending on where you actually live you may have it now. ;) Col

Papa_Bill
Papa_Bill

...to link your home telephones over your power wiring, but I agree, it probably wouldn't be suitable in a business environment.

b4real
b4real

Yes - saw that solution and I agree with your comments. But, to the root of the discussion, I believe the other end of that product is a copper networking interface!

LedLincoln
LedLincoln

Ethernet over water mains has proven to be too cumbersome for commercial viability. Maintaining the proper twist between the cold water pair and the hot water pair gives the typical plumber fits.

HAL 9000
HAL 9000

The available Spectrum. After all with what works OK now try pumping 1,000,000 users through the same frequency in a 1/2 square mile radius and see how well it doesn't work. WiFi is OK now because it's not yet saturated but it will get there just the same as Freeways are now Car Parks and no longer the free flowing traffic paths that they used to be. ;) Col

b4real
b4real

But, I doubt humans know everything about all metals and alloys. If copper gets expensive enough, we'll probably find a new alloy.

HAL 9000
HAL 9000

Copper is durable, Flexible, fairly Inert and cheap. Why would anyone try to replace it with anything. Well actually there is nothing that does the same job anywhere near as cheap so why bother even asking the original question? Col :^0

AnsuGisalas
AnsuGisalas

is steam-punk. Need'z be said more, eh, Schmott guy?

HAL 9000
HAL 9000

With 2 Fiber Cables running just outside my front door. But it all depends on where you like I could see the Fiber from the front door of one place 3 years ago but it was 200 meters away and no service from that. Just Half Speed Dial Up on a Split Copper Phone Line. Here the Telco's have a Fiber Backbone around the country which 2 years ago was approaching capacity so the Government took up an idea that I had suggested when they ran the Fiber of adding another cable for their own use. They saved 4 Million on the original Roll Out so the new fiber Cable is now costing only 4 Billion. They have decided on running Fiber to the Door now which is better but being a Government I'm sure that somehow they'll manage to stuff it all up. Currently the Opposition is complaining that it's too expensive and if they get elected they'll kill off the National Broadband Network as no one is going to pay for 100 MBS and at what they estimate to be 40 Billion it's way too expensive. However if they had of been in power when they did any National Infrastructure in the past it wouldn't have been Cost Effective to do so it wouldn't have got done. As there is now a Minority Government pushing the NBN it's unlikely to be finished which will be a pity, but as the Telstra legislation has passed, their Fiber Network is now part of the NBN which will speed up the roll out so I'm not really sure just how much will get done in the end. Currently the competition of the Proposed NBN has pushed my ISP to change their Data plans from 50 GIG per month to 1 TB per month which even I'll find hard to exceed. All of their new customers are getting the same plan but with a 100 MBS connection speed instead of their older 20 MBS which I'm on, so I suppose we'll see what happens. I don't think that the NBN will get finished but it's something that we really need and it's considered as so good that the Opposition Spokesmen on the issue who is using every excuse to bag it has just invested 1 Million of his own money in a company who's going to make lots of money out of the NBN so apparently even he can see the benefit although he bags it at every opportunity. :D Col

Oz_Media
Oz_Media

A fibreglas type of conudtor that was just blown in old serwer lines. Something to get started on in your bsement on Friday nights.

TobiF
TobiF

Whenever digging, they'll make sure to put DUCTS. Fiber can always be blown in afterwards!

Oz_Media
Oz_Media

They do it here too, but it is for potential future use. Carriers drop fibre when they get the opportunity, due to the exorbitant costs. The fibre here is generally used for telecom though and not offered as a residential data service.

Oz_Media
Oz_Media

Isn't your existing infrastructure really outdated though? Justifying the cost to the dmarc points.

AnsuGisalas
AnsuGisalas

I don't remember if it was the state or Stockholm municipality or whatever, but they made a policy decision: If there's road works, or sewer line work, or heating pipe works - and that street doesn't have fiber already - it gets fiber. They saved on bulk, obviously, and the ground is open anyway, so it cuts a lot of cost. Good policy. Not the kind that can be done with weak government.

Papa_Bill
Papa_Bill

...and Fanon and some others used to make a home intercom system that "worked" over AC lines. Extremely noisy especially if you had any florescent light at all.

Papa_Bill
Papa_Bill

much easier in fact than cat5. Even polarity isn't critical on phones, and bend angles aren't worth considering. Connectors are in the same class and use very similar tools. I considered trying one of the 'phone-over-AC systems, but the price didn't seem worth the effort.

b4real
b4real

I'd probably prefer wireless at home instead of the power line carrier.

CharlieSpencer
CharlieSpencer

I think a cable box or something similar was what I used one for back in the '90s.

Kevin@Quealy.net
Kevin@Quealy.net

It works pretty well except occasionally it'll get "stuck" and tie up the phone line until you reset it. It's easier than running a new cat 6 cable.

CharlieSpencer
CharlieSpencer

I tried one of these with no success. I admit it's been several years; the technology may well have improved. Frankly, I'd rather get under the house and run CAT-5. It can't be any more or less difficult than TV coax, and probably easier than telephone.

FAST!!!
FAST!!!

Very funny! Thank you.

LedLincoln
LedLincoln

Ten gig over gas has great potential, however.

b4real
b4real

Is this really a strategy?

Editor's Picks