Re: [HSSG] 40G MAC Rate Discussion
There is silicon available today that can run 4 x 10GBASE-KR as an
aggregated link (4 x 10). It has the advantage that it can support links
to multiple edge devices or a 40G aggregated link between redundant
switch fabrics. A single 40G link would be the same cost and would take
the same real estate as the 4 x 10 links. If you invest to make a new
generation that can support a single 40G MAC the added value will be
minuscule, I don't see how you could recoup the investment. On the other
hand, it would be a similar investment to develop the 100G MAC and such
a device is more suitable as an uplink for a switch fabric that is
serving 16 x 10G blades. The 100G MAC, along with the attached fabric
will be used for the inter-switch links and will be more immediately
useful, a 40G single MAC would be marginally more useful than the 4 x 10
you already have. I know where I would bet my investment dollars...
Ali Ghiasi wrote:
> Marcus and Others
> I like to present another point of view in support of 40 Gig MAC.
> We currently have the following option on the backplane side
> - KX-4 (XAUI) 10Gig
> - KR (1 lane ) 10Gig
> The natural next step for backplane Ethernet will be to operate KX-4
> lanes at 10.3125 Gbaud.
> Regardless of what decision we make in the HSSG 40Gig MAC will exist
> for the backplane.
> Assuming we will define the 40Gig MAC sooner or later then allowing 40
> Gig MAC
> for front panel becomes even more compelling, specially when 100Gig is
> overkill for these
> applications in near term. If we define 40Gig MAC in the HSSG then
> defining 40Gig
> backplane becomes travail.