[j-nsp] TCAM full on EX8200?

[Pavel Lunin]

> existing commodity chips which CAN do IP and some pretty deep hashing
> already,


This is where my doubts start :) You've mentioned QFX — is there any
evidence they are much smarter in hashing than EX?

Personally my take is that PTX missed the mark as far as interesting
> target customer size goes. There are still a *LOT* of places where you
> could very easily replace a $1mil core T/CRS box at $retardecarrier with
> a $10-20k 1-2U box that does 64x10GE, or 40GE, etc.


Completely agree. But IMHO (what I tried to point out in my posts above)
this requires either a feature compromise (weaker LB, no facility
protection, no ability to terminate inter-AS Option-C tunnels right into the
core, maybe something else) or reinventing a completely new ASICs
(simpler but not simple).

I believe, both ways are possible, really. Let's be honest, even smooth LB
can be thrown out, if you can use an $N×10k box to kick your competitor, who
uses $N×1M one. On the other hand, as you mentioned, there are not much
evidence of a big market for new good label-oriented ASICs. But isn't it so,
because no product means no market? Yes, OpenFlow can help to sort this out
:)


> Of course the large carriers with the most to gain also tend to be the
> least innovative and
> the most unwilling to consider cost when designing their
> architecture, until they go bankrupt of course. :)


Oh, yeah! This why the Juniper's bias towards top players can one day become
a banana peel.