x
Routing

Rumor: Cisco Prepping CRS-1 Successor

The next Cisco Systems Inc. (Nasdaq: CSCO) core router, the MSC 120, could be announced this month, according to Oppenheimer & Co. Inc. analyst Ittai Kidron.

In a note issued today, Kidron adds that the router, with up to 2.88Tbit/s of capacity by some metrics, probably won't ship until mid-2011 due to problems with the switching chipsets.

Kidron thinks the router will have a per-slot capacity of 120Gbit/s, which would appear to outdo the 100Gbit/s of per-slot capacity offered by the Juniper Networks Inc. (NYSE: JNPR) T1600 or the Alcatel-Lucent (NYSE: ALU) 7750 Service Router.

It would certainly best the CRS-1, which carries just 40Gbit/s per slot.

A 24-slot version of the router would therefore have 2.88Tbit/s capacity, although that figure doesn't address the switch-fabric capacity or the aggregate capacity of the router's interfaces.

Kidron also thinks a 12-slot version is in the works.

— Craig Matsumoto, West Coast Editor, Light Reading

Page 1 / 2   >   >>
dkh 12/5/2012 | 4:43:39 PM
re: Rumor: Cisco Prepping CRS-1 Successor

Just a couple of years late (give it 6-12 months more for the actual product delivery).


If you are curious, Huawei also claims they have a 100GE capable core router with a twist. According to their video, it is targeting the era of the "Ultra Boardband" )


http://market.huawei.com/hwgg/...


 

Pete Baldwin 12/5/2012 | 4:43:38 PM
re: Rumor: Cisco Prepping CRS-1 Successor

How about RHFR?

Pete Baldwin 12/5/2012 | 4:43:38 PM
re: Rumor: Cisco Prepping CRS-1 Successor

What I find interesting is the sudden importance of per-slot capacity, here and on the ASR 9000 ... versus the CRS-1's emphasis on building a massive fabric.


Cisco makes a big deal out of the fact that people ridiculed CRS-1's scale back in 2004.  But is this announcement an admission that the CRS-1 scales in the wrong dimension?

Stevery 12/5/2012 | 4:43:38 PM
re: Rumor: Cisco Prepping CRS-1 Successor

We've gone from the GSR to the BFR to the HFR.  What's next?

forsythe.meadows 12/5/2012 | 4:43:37 PM
re: Rumor: Cisco Prepping CRS-1 Successor

huawei will likely announce in 2010- they really did something beyond off the shelf this time


and they will fix their spelllllling errrrrrors


they indeed grabbed an asic team that will outwit their off-the-shelf adventures


whether they will enjoy much core is not yet certain! too new for a box that sits in the "oops - we lost new england" category


cisco is late to the bash - 2010 is the year of the demo!


 

dkh 12/5/2012 | 4:43:36 PM
re: Rumor: Cisco Prepping CRS-1 Successor

> What I find interesting is the sudden importance of per-slot capacity, here and on the ASR 9000 ... versus the CRS-1's emphasis on building a massive fabric.


Craig,


Let's not confuse creative marketing with revenue-generating product capabilities.


ASR9K started life with 40G/slot linecards and only now starts gaining 80G/slot blades. CRS-1 fabric does only 40G/slot, but can (theoretically) support many destinations for multichassis.


Practically scaling a multichassis beyond 8xLCC involves solving too many software, hardware and systest issues. That's why it takes several years between shipping a single-chasis and multi-chassis system, in which timeframe the newer generation of silicon becomes available rendering the old multichassis system useless. For analogy, imagine someone offering a room full of interconnected Cisco 7500 routers - not a good investment from space, energy and throughput perspective.



We all heard senior Cisco executives dropping "petabytes" here and there and I think Guinness record book still lists CRS-1 as a "92Tbps router", but this, of course, is pure nonsense for technically handicapped cat owners (somehow dog owners tend to know better).



So, technically, upping the ante for CRS-1 slot capacity is a significant achievement.


What's much less impressive is timing - Juniper T1600 has been shipping since 2007, which means that Cisco's new box risks becoming obsolete immediately after the first product shipments (2011?)





schlettie 12/5/2012 | 4:43:34 PM
re: Rumor: Cisco Prepping CRS-1 Successor

No one with any sense releases a new chassis without a plan to scale the slot capacity 3-4x in the future.  Cisco didn't forsee 100 GE?  I wonder if they will hatch yet another OS for the new box.


Juniper stands to pick up a lot of business from stranded CRS-1 customers.

TomS6 12/5/2012 | 4:43:34 PM
re: Rumor: Cisco Prepping CRS-1 Successor

note to self: don't pay for any Oppenheimer & Co reports on networking gear.


 


Abandoning the CRS-1 as the article/rumor implies would be beyond stupid and won't happen. Maybe a little research on the CRS-1 (like the parts list) would be enlightening for the author.

Pete Baldwin 12/5/2012 | 4:43:33 PM
re: Rumor: Cisco Prepping CRS-1 Successor

Schlettie:


>I wonder if they will hatch yet another OS for the new box.


Oooh, good one.  There's no way they'd bring in yet another OS, right? Right?


Pete Baldwin 12/5/2012 | 4:43:33 PM
re: Rumor: Cisco Prepping CRS-1 Successor

Dkh -- Great points, thanks for the post.  Actually, it was the marketing I was talking about -- interesting to see which ultra-big dimension becomes important from year to year.  I doubt we'll hear about "72-box multichassis" this time around, maybe for the exact reasons of practicality that you mention.


>What's much less impressive is timing - Juniper T1600 has been shipping since 2007, which means that Cisco's new box risks becoming obsolete immediately after the first product shipments (2011?)


That's a good point!


Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE