2

In my work, I come across with a cable with TB/s capacity and speed in GB/s.

I did some search and found and article from Cisco trying to explain the two terminologies in terms of a water tap. The capacity is like the width of the tap pipe and mouth; while speed is how faster water is flowing through the pipe.

I understand this example, but I found it hard to apply to the cable situation, and eventually hard to understand the two terms.

In the water tap example the capacity is measured with a unit without per second. But in the world of cable, capacity is measured TB/s.

The definition of bandwidth is how much data can be transferred in a second. Since capacity say this cable can support 1TB/s, doesn't it mean given an unrealistic ideal world where there are no lost, interference whatever, it can allows 1TB of data to pass through? If so, isn't this already the capability of this cable. How should I understand the term speed and why should the cable states it?

Reference: https://learningnetwork.cisco.com/blogs/vip-perspectives/2015/10/16/bandwidth-vs-speed

3

Your question is pretty confusing, and it is somewhat misleading, but I will try to clarify.

The capacity to which you refer is called the bandwidth, and it is measured in bps (bits per second). For example, 100 Mbps. The speed of transfer on a cable is fixed and limited by physics (the speed of light in the medium of the cable), and it it relatively the same for all your cables. For example, the speed of light in copper cable is about two-thirds the speed of light in a vacuum. The bandwidth is more a function of the device interfaces than it is of the cable.

There are cable standards set by ANSI/TIA/EIA and ISO/IEC. To facilitate the bandwidth of the device interfaces, the cable must meet some parameters. This can get very technical and complicated, which is why the standards bodies have created various cable standards. For example, ANSI/TIA/EIA has categories for copper cabling, and ISO/IEC has cable classes. The various standards define parameters like Insertion Loss, NEXT, FEXT, Return Loss, Propagation Delay, Skew, etc. Depending on the particular set of parameters a cable has, the cable is rated for a maximum frequency it can transmit, e.g. 100 MHz for Category-5e. How the interfaces encode and signal on the cable determines the bandwidth, but a cable must meet the requirements of the interfaces in order to function at the bandwidth of the interfaces.

A big part of whether or not the cable can function correctly at a particular bandwidth is determined by the cable installation. There are standards for this too. For example, ANSI/TIA/EIA 568, Commercial Building Telecommunications Cabling Standard. Poorly installed cable will not function correctly. All components of a cable path (cabling connectors, etc.) must be rated the same, installed properly, and tested with expensive equipment to validate that they perform correctly.

An example of the cable bandwidth would be Category-5e cable. If it is properly installed, the cable can work at 10BASE-T (10 Mbps ethernet), 100BASE-TX (100 Mbps ethernet), and 1000BASE-T (1 Gbps ethernet), but not at 10GBASE-T (10 Gbps ethernet).

It is the interfaces of the devices to which the cable connects that determines the bandwidth of the link. For example, The maximum bandwidth on Category-5e cable would be 1 Gbps. If you try to use it with devices that only work at 10 Gbps, then it will not work at all. Some people may think that the cable will transmit at 1 Gbps in this case, but it doesn't work that way. The interfaces on the devices will send data at frequencies that the cable simply cannot reliably handle, and you will receive garbage at the other end. This is where the comparison to a water pipe fails.

  • Thanks for the detail explanation. It helps me understand the topic a lot. Back to the question, first I shouldn't let the water tap example bothers my mind too much because it does not help me in understanding the two terms. Second, when I am evaluating this cable I should be aware of its capacity because it describes given a correct installation, it can transfer that amount of data per second; And finally for the speed, it is just a matter of the cable material, can I consider it as the quality of my cable? And I shouldn't care it too much, given it meets the average speed of that material. – Calvin Lau Jul 13 '17 at 16:01
  • There is a lot of crap sold for copper cabling, and manufacturers find it harder to meet the higher categories, so they start doing things like putting shield on, and that complicates things because the shield must be continuous and grounded, at least, on both ends, and the equipment must support it. You need to get cable of at least a category that matches your equipment. Get it professionally installed and tested. Pulling a horizontal cable too hard, or bending it too tightly, even once, can permanently ruin the cable. – Ron Maupin Jul 13 '17 at 16:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.