Why is everyone still using CAT 5? - Page 2

I just finished building a network for the largest non-profit organization in the world. The project cost 700 million dollars, and this organization wound up with more money in the bank after they spent 700 million on this building than before it.

Point im making is, they have deeeeeeep pockets and you know what they are sending to the desktop? 10/100 switched. They are running 10 gig in the backbone, but 100MB to the desk.

Once again, cable is not the problem. It never is. Its the limitation of the electronics.

I dont think the price of a NIC card that can support 10Gig will ever be possible. If it is, it will require SM fiber.

Besides, right now, a Berk-Tek LANMark 10G or a Systimax 10/2091 cable is about 6 times the cost of a 2-strand SM zipcord. And i dont see it ever being cheaper.

As for consumer use, come on folks, admit it, 10 gig in your house? Maybe in 15-20 years........

"640K ought to be enough for anybody. " Bill Gates, 1981



OK, OK... so it's not a real quote, but you get my point.

First off, what you are running at a non-profit, where the people working on the computers are probably running basic office applications, differs from what you might want in your home in 10 years. I'm not saying the applications are there yet, but we can hope they will be. And when/if the applications exceed 1GE in the home, the non profit computers will still be running office programs. Also, you are talking what, hundreds of thousands of feet of cable, such that an incremental cable cost differential will make a big difference to the bottom line. This is also different than in the home.

Since 10GE is not spec'ed at all over cat5e (although it might work for short distances) the question boils down to: Will you ever need > 1G in your house? The answer is unclear, but if an extra (small) cost differential and some careful wire routing of cat6 will make it possible, then I figure it's worth it.

Realize also that 10G is a different beast than 1G. If the line is too noisy, you get nothing. It's not like it can slow down to 5G, or 3G if the line is noisy.

On the other hand, there are discussions in the IEEE to try to create a more multi-tiered approach, so that might save your cat5e installation, making it possible to run 3G or 5G over it even if it is too noisy for 10G. If that happens it might save your cat5e installation at the cost of some (probably unneeded) bandwidth. In that case you will be right about just needing different electronics :-)

Cost-wise, 10G is expected to follow 10/100 and 1G. Initially it will be $hundreds per port, and used only in the data center where the cost savings over fiber are high. Quickly after that, it will move into high end servers (still for $hundreds per port) to better connect them to a backbone network. Thus it will probably be possible (although unnecessary) to buy a 10G PCIe card for your computer in a year or two, for a bunch of money. As volumes ramp, and manufacturing costs lessen, it could easily filter into desktop cost levels, although with the hard-to-meet cable specs, and the lack of applications requiring > 1GE it might not gain the acceptance that 10/100 and 1GE has.

This is all my personal opinion, btw, and I am not speaking for my company, or for the IEEE.

The reason to move to a higher bandwidth cable is that in the future we'll be moving this stuff around our homes in data packets, not line-level. The MHz that the cable makes available is interesting, but the most relevant is what data bandwidth can it support. Putting Cat5e/6 is good future proofing even if you're just using it for baluns or another analog distribution method today, because tomorrow you're going to use those lines for data distribution.

ok, ill buy that. going on the assumption that one day we will be watching IPTV, everyone needs 10gig ability in their homes.

anyone who wants to buy 'reportedly' 10 gig capable cabling, send me a pm. ill sell it to you!

I think the answer is simple. It costs more. From a more pragmatic standpoint, all equipment that uses UTP has converged on Cat5, so all standards committees understand that in order to sell anything in the future, their technology has to work over Cat5. The cost for a person or company to run new cable for a single application would be prohibitive most times since fiber is out there for higher speeds already. So what you have is chicken/egg. You need the compelling app before investment, but the compelling app won't want to be on an island without sales, so they'll work hard to retrofit it to Cat5 or move it to fiber. Simple as that. HDTV distribution over IP works great on Cat5, what compelling home application do you foresee that will tap Cat5?

It's not unlike how long it has taken us to move the Internet to an IP optimized infrastructure. A good majority of the Internet still travels over the network built for the telephone. Other arguments you could make would be that we should all stop installing these stovepipe home automation, security, audio etc., systems. Don't you all see that the future will have all of those systems running over IP and fully interoperable? Those systems have already moved toward Cat5, IP migration is next. Why install Coax...UTP is the future. Unfortunately, the future is not quite here yet, so we install what works and lives in a price competitive realm realizing like all technology, we'll have to upgrade it later.

Don't get me wrong, I'm one of those "first guy on the block" at least in terms of home networking, but I didn't hesitate to grab Cat5e when I saw the price difference and the general product availability. I have only one suggestion...don't skimp on Cat5 runs...install it everywhere. Wireless is great, but it will always be behind Cat5 as Cat5 will always be behind fiber. The big unknown to over-engineer for is the new IP based Cat5 applications that haven't been implemented/dreamed up yet. If you install it, you will use it. I have 4 Cat5s (data only) running to my theater area. I can already see it being tapped (MCE, Receiver, Squeezebox, TV) - all over IP.