Speed and rate of data transfer

  • #26
555
9
Back to the original question, it will be quite impossible to send full CMOS-level (5V) signals down 100 meters of any reasonable wire at 1 Gbps. You need a SERDES (serializer-deserializer) to make this happen. Here is a good place to start if you're actually interested in digital communications:

http://www.xilinx.com/publications/archives/books/serialio.pdf

I suspect this was just a "theoretical" question. If so, you need to think about the concept of latency. That is, the information can still get from point a to point b but it doesn't have to be instantaneous. Thinking about electrons in a wire is usually not the most useful way to think about a digital communications link. Thinking about EM waves is usually far more fruitful.

As an example of latency, we can send signals to the rover on Mars but there is a 12 minute (or so) latency (depending on orbit). But the information still gets through. How is that possible?
 
  • #27
sophiecentaur
Science Advisor
Gold Member
24,797
4,614
Back to the original question, it will be quite impossible to send full CMOS-level (5V) signals down 100 meters of any reasonable wire at 1 Gbps. You need a SERDES (serializer-deserializer) to make this happen. Here is a good place to start if you're actually interested in digital communications:

http://www.xilinx.com/publications/archives/books/serialio.pdf

I suspect this was just a "theoretical" question. If so, you need to think about the concept of latency. That is, the information can still get from point a to point b but it doesn't have to be instantaneous. Thinking about electrons in a wire is usually not the most useful way to think about a digital communications link. Thinking about EM waves is usually far more fruitful.

As an example of latency, we can send signals to the rover on Mars but there is a 12 minute (or so) latency (depending on orbit). But the information still gets through. How is that possible?
In fact, the black box approach is usually the most successful in engineering. You do the link budget which tells you your likely carrier to noise ratio. This will tell you the likely error statistics for your choice of coding and modulation and you then decide whether it's a pass/fail.

Life is too short to think in terms of studying every link in every chain.
 
  • #28
555
9
In fact, the black box approach is usually the most successful in engineering. You do the link budget which tells you your likely carrier to noise ratio. This will tell you the likely error statistics for your choice of coding and modulation and you then decide whether it's a pass/fail.

Life is too short to think in terms of studying every link in every chain.
How true that is. Start with an ideal model and only break open the model as each point fails. Always good advice.

I would say that in a serial link you would be well advised to think more in terms of eye opening rather than C/N. In a serial link it is usually dispersion that kills you, not noise.

Still, if the goal is "understanding" in some sense, then EM waves are a better paradigm than discrete electrons.
 
  • #29
sophiecentaur
Science Advisor
Gold Member
24,797
4,614
Electrons are probably the worst possible introduction to 'electricity'. Never touched them in my early electrical learning. Nowadays they are used by people who have very little understanding of Science to confuse kids with their teaching. And it's not even their fault.
 
  • #30
755
78
On the topic of latency of data communications links here are some stats you may find interesting:

At 100 gigabit Ethernet, the spacing between bits on the fiber optic cable is a mere 2mm.
A long reach (40Km) 100GBE link will contain 20 million bits in flight.
Corning Glass has shipped over 30 million Km of long haul (LEAF) fiber.

Thus, at any frozen moment in time, this corresponds to 15 terabits of data in flight within these cables.
 
Last edited:
  • #31
sophiecentaur
Science Advisor
Gold Member
24,797
4,614
On the topic of latency of data communications links here are some stats you may find interesting:

At 100 gigabit Ethernet, the spacing between bits on the fiber optic cable is a mere 2mm.
A long reach (40Km) 100GBE link will contain 20 million bits in flight.
Corning Glass has shipped over 30 million Km of long haul (LEAF) fiber.

Thus, at any frozen moment in time, this corresponds to 15 terabits of data in flight within these cables.
There is a directly similar acoustic situation. When you shout at someone in the distance, your whole sentence may be hanging in the air, on its way to them after you close your mouth finally.
Latency is only a problem when synchronism is important. This can be on a circuit board where line lengths on a Bus need to be equalised to within a mm or two.
 

Related Threads on Speed and rate of data transfer

  • Last Post
Replies
2
Views
2K
  • Last Post
Replies
12
Views
3K
  • Last Post
Replies
1
Views
3K
  • Last Post
Replies
3
Views
928
  • Last Post
Replies
3
Views
1K
Replies
11
Views
7K
Replies
2
Views
1K
Replies
2
Views
5K
Replies
11
Views
744
Top