[Cialug] Masters and Slaves

[Colin Burnett]

On Sun, Nov 2, 2008 at 12:48 PM, Zachary Kotlarek <zach at kotlarek.com> wrote:
>
> I was reacting to the "no arbitration/addresses" part of your note --
> there's still a single physical link that uses addresses to distinguish
> between multiple endpoints, and there's still the potential for multiple
> endpoints to generate more data than the single uplink can transmit, which
> demands arbitration. Those two functions may not happen as part of the
> fundamental signaling protocol -- they may be encapsulated as data across an
> unarbitrated, addressless link that only the HBA and port multiplier see --
> but they're still part of the overall SATA system.

Yeah, still different layers.  You're talking "IP address" and I'm
talking "MAC address".  Ethernet doesn't care about IPs and however
the ends interpret that IP is irrelevant to ethernet frame routing.

PATA needs to know master/slave to know which is being talked to
because they're both listening and both could be talking.  SATA
doesn't need a "MAC address" because there is no ambiguity on the
line.  The concept of a port is strictly a higher layer function.

Same words, same meanings, but different layers. :)


It is interesting to note, though, the move away from shared bus to
serial streams.  SATA, PCIe, and even ethernet in the last years (hub
vs. switch).  Does anyone make a gigabit *hub*?  SATA covers device to
mobo; PCIe covers intra-mobo; and switched ethernet covers
inter-computer.

I believe firewire and USB are also P2P with addressing done at a
higher layer, yes?  Both are bidirectional serial streams.

It seems the industry has all but abandoned multiple access and
parallel.  I suspect, however, the cycle will come back around and
people will realize that you could reduce hardware by sharing wires
and get 8 times the data if you send it on 8 different wires!  Wow!
:)


Colin