[Gllug] re: SCSI vs Serial ATA

[Jack Bertram]

* Andrew Halliwell <ah at gnd.com> [040614 12:23]:
> And verily, didst t.clarke announce to the hordes:
> > If electricity flows at about 1 foot a nanosecond, the delay in one strand of
> > a signal over another would seem to be absolutely minimal as a percentage of
> > the clock-rates currently in use.   Would not a 'bit width' on a parallel bus
> > twice that of on a serial bus, with a smaller 'sampling window' in the middle
> > of the bit to cope with timing differences still achive an overall bandwidth
> > 4 times greater than a serial bus.   Or have I missed something in the
> > arguments??
> 
> It's a little more complicated than that when it comes to high speed data.
> With parallel, not only do you have the speed of the electricity travelling
> down the wires,but delays caused by the internal resistance across the
> wires, impedance across the wires and capacitance between the wires and
> inductance between the wires causing interference...
> 
> These lead to rather complex physical maths equations, and if one strand has
> a slightly higher resistance and impedance than the others...

Indeed.  And at slow data rates, these complications aren't too
important (hence parallel printer cables being faster than RS232).  But
for any serious bandwidth, timing becomes so important that it's a
really big issue.

j
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mailman.lug.org.uk/pipermail/gllug/attachments/20040614/dc651eb3/attachment.pgp>
-------------- next part --------------
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug