[j-nsp] ACX50xx l2circuit counters

Saku Ytti saku at ytti.fi
Tue Jun 21 07:31:38 EDT 2016


On 21 June 2016 at 14:23, Mark Tinka <mark.tinka at seacom.mu> wrote:

Hey,

> Take Policy Map for the MX, for example. I worked with Juniper for 2
> years to develop this feature, and was lucky it could be done on the
> Trio because of that design. They can't do it on any other M-series, for
> instance, and certainly not the DPC's which are external chips. It's not
> a feature they foresaw when developing the Trio, but they managed to
> make it work without sacrificing performance after-the-fact.

You could do it in every NPU, NPU is run-to-completion, you give it
real software to run, and it does something. You're only limited by
time.

All IP/L3 stuff is done on internal I-Chip, evolution from IP2.
EZchip2 is there just for L2, QoS or something like that. Ironically,
the EZchip2 itself could have done it for sure, but perhaps not at an
acceptable rate.

> When I started testing the ME3600X/3800X in 2009, the Nile chip did not
> support QPPB. After working with Cisco for 6 months, I managed to
> convince them to re-spin the ASIC before they started manufacturing the
> box for general shipping. But such opportunities are few and far between
> (basically, non-existent), and I wouldn't bank on them being the norm,
> ever. Which is why flexibility in the future is easier with an in-house
> chip, even if it's not always guaranteed.

This is absolutely fantastic and BRCM would probably require much
larger volume to respin anything.

-- 
  ++ytti


More information about the juniper-nsp mailing list