Hi Ben, A TCAM can have an entry which is broader than 64 bits.
An optimization could be to have some entries 64 bits and some 128 bits to be able to more effectively use the TCAM memory based on the usage. Thanks, Vishwas On Mon, Nov 8, 2010 at 10:23 AM, Hemant Singh (shemant) <shem...@cisco.com> wrote: > Oh, I am still on US EST time zone and waking up in the middle of the night > in Beijing. Please see in line below. > > -----Original Message----- > From: ipv6-boun...@ietf.org [mailto:ipv6-boun...@ietf.org] On Behalf Of Ben > Jencks > Sent: Monday, November 08, 2010 5:56 AM > To: Thomas Narten > Cc: ipv6@ietf.org > Subject: Re: draft-ietf-6man-prefixlen-p2p-00.txt > > >>I'd like to hear from a router vendor on the impact of this on TCAM >>design. I'm under the (possibly mistaken) impression that frequently >>only the first 64 bits are allocated space in the TCAM. If you need to >>route on all 128 bits in hardware, it could halve the number of routes >>you can carry. > >>This has probably already been addressed, but I couldn't find it in >>the archives. > > Speaking as a router vendor, the possibly-mistaken-impression is so not the > way IPv6 high-end routers silicon works when the router uses a TCAM. Sorry, > I cannot provide any more details on our router implementation, but I can > assure this audience that Thomas' belief is correct that IPv6 routing > continues to be based on CIDR. > > Hemant > -------------------------------------------------------------------- > IETF IPv6 working group mailing list > ipv6@ietf.org > Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 > -------------------------------------------------------------------- > -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
