[187100] in North American Network Operators' Group
Re: New Switches with Broadcom StrataDNX
daemon@ATHENA.MIT.EDU (Phil Bedard)
Tue Jan 19 17:17:51 2016
X-Original-To: nanog@nanog.org
Date: Tue, 19 Jan 2016 17:17:44 -0500
From: Phil Bedard <bedard.phil@gmail.com>
To: Jeff Tantsura <jeff.tantsura@ericsson.com>,
Colton Conor <colton.conor@gmail.com>
In-Reply-To: <E0124EA5-2234-4957-AF7D-550CAFDDC246@ericsson.com>
Cc: NANOG <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
Good point, there are many people looking at what I call FIB optimization r=
ight now. The key is having the programmability on the device to make it ha=
ppen. Juniper/Cisco support it using policies to filter RIB->FIB and I beli=
eve both also do per-NPU/PFE localized FIBs now. I am not sure if that=E2=80=99s s=
omething supported on this new Broadcom chipset. Depends on your network of=
course and where you are looking to position the router. =20
Phil=20
-----Original Message-----
From: Jeff Tantsura <jeff.tantsura@ericsson.com>
Date: Tuesday, January 19, 2016 at 11:46
To: Colton Conor <colton.conor@gmail.com>, Phil B <bedard.phil@gmail.com>
Cc: NANOG <nanog@nanog.org>
Subject: Re: New Switches with Broadcom StrataDNX
>Hi,
>
>Some points:
>1.DNX SDK is significantly different from SGX, adopted by Cumulus and such=
, yet to be done, and this is not negligible amount of work
>2.if you are not interested in capacity but in scale, there=E2=80=99re other BCM=
chips, perhaps more suitable
>3.you don=E2=80=99t have to have all the forwarding entries populated in silicon=
, as an example - take a look at http://sdn-internet-router-sir.readthedocs.=
org, code at https://github.com/dbarrosop/sir, one could also leverage appro=
ach we have taken in EVPN - decoupling RIB from FIB completely
>4.NG silicon will do 1M+ LPM's
>
>Cheers,
>Jeff
>
>
>
>
>
>
>
>On 1/19/16, 06:29, "NANOG on behalf of Colton Conor" <nanog-bounces@nanog.=
org on behalf of colton.conor@gmail.com> wrote:
>
>>I was hoping this new Broadcom chip would be able to support enough route=
s
>>to hold a full BGP table, and be used for something like cumulus linux. I
>>have no need for 100G, but 10G and 40G on a platform with deeper buffers
>>sounds nice.
>>
>>On Tue, Jan 19, 2016 at 1:01 AM, Phil Bedard <bedard.phil@gmail.com> wrot=
e:
>>
>>> The BCM88670 (Jericho) is what powers the new Cisco NCS55XX devices. Th=
e
>>> processor is linerate above around 100 bytes per packet without externa=
l
>>> TCAM, supports 256K IPv4/64K IPv6 FIB entries (or mixed amounts). Thes=
e
>>> chips are being used for high scale 100G, the initial NCS5508 linecard =
is a
>>> 36x100G QSFP28 one.
>>>
>>> Juniper has chosen to use their own silicon for most of their dense 100=
G
>>> platforms, but you=E2=80=99ll see these chips used by pretty much everyone el=
se I
>>> imagine at some point in the next year.
>>>
>>>
>>>
>>> Phil
>>>
>>> -----Original Message-----
>>> From: NANOG <nanog-bounces@nanog.org> on behalf of Colton Conor <
>>> colton.conor@gmail.com>
>>> Date: Sunday, January 17, 2016 at 18:15
>>> To: NANOG <nanog@nanog.org>
>>> Subject: New Switches with Broadcom StrataDNX
>>>
>>> >Does anyone know when the switching and router vendors will release th=
eir
>>> >new models with the Broadcom BCM88370 and BCM88670 chips? It looks lik=
e
>>> >these chips could be used as a carrier grade router and/or metro E dev=
ice.
>>> >
>>> >More information here:
>>> http://www.broadcom.com/press/release.php?id=3Ds902223
>>> >
>>> >and here:
>>> >
>>> http://www.nextplatform.com/2015/03/19/new-dune-chips-enable-heftier-sw=
itches/
>>>
