Athlon XP: Awaiting 64-bit Architecture
On April 22, AMD announced a server version of its 64-bit Opteron processor. Unfortunately, the server version was announced. In addition, most likely at the beginning the 64-bit platform will…

Continue reading →

I love the heat at the beginning of the year!
Finally, winter has come. The time has come when it is necessary to seriously think about cooling computers. In magazines, the topic of computer cooling is raised, as a rule,…

Continue reading →

Metrology notes
Despite the variety of generally accepted tests, recently their results have often added confusion than to clarify the essence of the phenomena. There are tests that allow you to measure…

Continue reading →

We overcome the limitations

Standards for SCS based on twisted pairs for Ethernet 10Base-T and Fast Ethernet 100Base-TX networks stipulate that:

the maximum length of horizontal wiring is 90 m;
the total length of the connecting cords on the cross and in the working area should not exceed 10 m.
In most cases, these restrictions fully satisfy the LAN needs of a small office. However, quite often you may encounter the fact that one of the LAN workstations needs to be placed at a distance of, say, 150 meters from active equipment. In this case, if you adhere to the standards, it is necessary to install additional cross and active equipment (hub or switch) at a distance of up to 90 m from the main equipment and extend a line from it to the workstation or change the signal transmission medium, for example, to optical fiber. For a small office, such a solution is quite expensive and inconvenient, since it is not always possible to install additional active equipment in the right place, and changing the signal transmission medium entails the use of special devices – medium converters, which again comes down to significant additional costs.

Is there an inexpensive solution to this situation? Consider where the 90 m horizontal wiring for twisted pair cable came from. Standards do not arise from scratch; they are based on the capabilities of existing technologies and the needs of practice. A significant role in the adoption of current standards was played by the technology of manufacturing cables from twisted pairs of conductors. Recall how the transition from Ethernet to Fast Ethernet networks took place. Manufacturers of cable products and network equipment offered various options for building Fast Ethernet networks. Let us recall the 100Base-T4 or 100VG-AnyLAN technologies that have not found wide application, at least for us. In these technologies, the transmission of signals at a speed of 100 MB / s was carried out not in two, but in four pairs, and a category 3 cable could be used as a transmission medium. These standards arose at a time when category 5 cable was quite expensive, the number of installations on category 3 cables was quite large, and the practical needs dictated the use of high-speed network applications. The manufacturers were faced with the task of proposing a solution that allows data transfer 10 times faster, while not reducing the length of the segment, which for 90Base-T networks is 90 m. The technology for the production of Category 5 cables and active equipment was adapted just to these requirements.

More recently, while developing standards for gigabit data transfer technologies, manufacturers began to offer cables with improved frequency properties, which began to be positioned as cables of categories 5e, 6 and 7. The electrical characteristics of these cables significantly exceed the similar characteristics of category 5 cables.

Figure shows the attenuation of the signal (Attenuation) and crosstalk (NEXT) versus frequency for cables of categories 5 and 6 (segment length 100 m). We see that Category 6 cable has a significant (up to 25 dB) ACR value at a frequency of 100 MHz. Class D cabling standards normalize this value to about 4 … 10 dB (and different standards differ in different ways). This means that if we use Category 6 cable instead of Category 5 cable in a segment critical in length, then with a length of 100 m we will have an ACR margin of up to 15 dB. We can use up this extra margin by increasing the total segment length to some limits. With an increase in the segment length over 100 m and a fixed frequency (for example, 100 MHz), the level of the useful signal will decrease, and the level of crosstalk will increase; as a result, ACR = NEXT – A, i.e. signal to noise ratio is getting worse. An unreasonable increase in segment length can lead to a negative ACR value and even complete network failure. How much can a segment be extended to maintain the ACR required by standards? This mainly depends on the actual characteristics of the cable used and on the active equipment used. We can share our own results of a practical verification of the above considerations. Using a Category 6 cable PiMF (300 MHz) in the Fast Ethernet network segment, we achieved stable network operation with a segment length of 150 m. Using Category 5e cable in Ethernet 10Base-T networks gave a positive result with a segment length of up to 180 m.

Forecast for the Development of Computer Hardware and Software for 2001
Processors The most significant event is the appearance of two 64-bit processors (Intel and AMD) with different structures. In 2001, 64-bit processors still did not affect the market for processors…

...

Custom solutions for local networks of small offices
From the school bench, we are all familiar with some of the laws of nature, which are therefore called the laws that cannot be violated. However, in addition to laws,…

...

What is Athlon 64?
First of all, it must be emphasized that in the AMD64 architecture, the main attention was paid to the issue of compatibility with 32-bit applications that are widespread now. The…

...

New from Japan
The article was written by a former employee of the EPOS company Sergey Tkachenko. Currently, Sergey is studying and working in Japan, he is a graduate student at Chubu Daigaku…

...