[Arm-netbook] EOMA server standard

[Gordan Bobic]

On 10/24/2012 02:14 PM, luke.leighton wrote:
> On Wed, Oct 24, 2012 at 1:23 PM, Gordan Bobic<gordan at bobich.net>  wrote:
>
>> Possibly the best option might be a single 10G fibre port with
>> additional 4-5 copper gigabit ports for people who aren't set up for fibre.
>
>   my only concern is: standards *have* to be based around
> "non-optional-itis".  or, that whatever you put down it can have
> auto-negotiation.

Why does the number of external ports on the chassis have to be 
standardised? What is there to be gained from this? Normal servers don't 
have a "standard" number of network ports.

>   basically, if you have fibre and also 4-5 copper gigabit ports, all
> server card implementations *MUST* support fibre and also 4-5 copper
> gigabit ports.  even if you have fibre and only 1 copper gigabit port,
> all server cards *MUST* support *BOTH* interfaces, even if the I/O
> boards don't use the fibre or don't use the copper.

That's not something you'll put on the server EOMA card - that's 
something you'll put on the _chassis_.

I cannot imagine there being any point in having 10G ethernet directly 
on the EOMA card.

>   you *CANNOT* have "optional" ports on the server card.  you just
> can't - it causes complete chaos.

We appear to have had a misunderstanding on what we are discussing here. 
I thought we were discussing possible server chassis standards. I really 
don't think there's any point in even contemplating 10G ethernet on the 
EOMA card itself. We're probably 2-3 SoC generations behind that being 
even remotely useful - and by then you can devise a new, more advanced 
standard for the more advanced SoCs when they are actually available.

>   this is why i was looking at 10GBase-T because you can use the
> *exact* same 8 wires for 10/100/1000/10000 ethernet, it's all
> auto-negotiated.

Except the network port will suck up 5x more Watts than the whole of the 
rest of the card put together. I really don't see it as particularly 
desirable. :)

>   setting a lowest-common-denominator standard (where its sub-standards
> all have auto-negotiation), in a market that's about to expand, is
> very very tricky!
>
>   this needs a lot of careful thought, or just to say "nuts to it" and
> set a bar and tough luck for any SoCs that are below that level.

Since there are no SoCs that can do 10G on the chip, it's pretty 
academic. And given the sort of power envelope we are looking at, the 
chances are that it will never actually be workable with the copper variant.

Gordan