Speed, hardware or protocol?

**TcM** · 07-21-2008, 07:11 AM

Is speed (of anything like speed of PCI port) is a matter of protocol, or how the hardware is setup or it's a physical limitation in the hardware?

**MeTh0Dz** · 07-21-2008, 10:25 AM

The highest possible speed is determined by hardware. But the software that runs on the ports, drivers, etc. factor into the speed to which the hardware actually operates. Things like overhead and calculation time will decrease the actual speed that it runs at.

**TcM** · 07-21-2008, 10:54 AM

So, ok. My router and laptop are capable of communicating at 54Mbits (my router is capable to 108Mbits, but anyways)

Why is it that only in theory they should? In practice they never do it... I mean if it's hardware related, and it is capable of running at such speeds.. why don't they?

**MeTh0Dz** · 07-21-2008, 10:57 AM

Look up... The difference between Bandwith, Throughput, and Goodput. That explains why theoretical speeds are never reached.

**TcM** · 07-21-2008, 11:27 AM

That was just an example, I know something about those...

In many cases, the protocol states a speed which is not realistic... why?!?!

**MeTh0Dz** · 07-21-2008, 11:37 AM

If the protocol says that speed is possible, then it means that it can operate at that speed. However hardware has to have the ability to perform at that level.

**TcM** · 07-21-2008, 11:40 AM

But if in practice it never works at that speed, how can they jump to these conclusions?!

**MeTh0Dz** · 07-21-2008, 11:43 AM

They can create an enviornment in which these speeds can be reached. An extremely optimized enviornment with top of the line equipment that is designed to run such a protocol can reach these speeds.

These are just the speeds and performance markers that they have been able to max out at.

I think that is what it means. If not....

What they are saying is that in theory this protocol can operate at this speed, however current hardware/software technology will not allow such a mark to be met.