[arin-ppml] debunking the myth that Moore's law helps

In a message written on Fri, Dec 18, 2009 at 09:25:18PM -0800, Michel Py wrote:
> So, the switch part works at wire speed; that was somehow expected,
> given that the backplane capacity on a 4-port switch is a non-issue. The
> NAT part goes over 100Mbps but not much. Does everyone agree that a $250
> device with something like an Atom CPU would get ~800Mps NAT (if there
> was a market for that device today)?

I think this thread has gone slightly astray.

The initial comparison was of a "ISP grade" router in 1991, to an
"ISP Grade" router in 2009.  There is no firm definition of this,
so I can't put my finger on a single definition that would include
or exclude boxes.

That said, the "home routers" being discussed are not something a
business needing GigE service would purchase.  They are not something
an ISP offering managed CPE to a GigE customer would purchase.  Be
it filtering, routing protocols, command line management, ssh, there
are a host of features missing.

I think it is quite cool that these low cost, specialized application
devices exist.  I'm not sure how this furthers the discussion of
device growth related to bandwidth over time.

       Leo Bicknell - bicknell at - CCIE 3440
        PGP keys at
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 826 bytes
Desc: not available
URL: <>