x
100G Ethernet

100G Standards Aim for Lower Costs

Vendors are hopeful that the standardization efforts around 100-Gbit/s transmission will pay off in the form of lower costs.

The hope is that they can learn from the hurried 40-Gbit/s generation, which got its start during the bubble and led to a panoply of modulation formats. In turn, that prevented the industry from building up any economies of scale -- a factor that's coming into play now, as 40-Gbit/s deployments gain steam.

Analyst Danny Dicks outlines the situation in the recent Heavy Reading Insider report, "Optical Transport: 40G/100G Interest Shifts Into Overdrive."

"While 2008 and 2009 have seen deployments of 40G systems around the world, such systems are not yet challenging 10G in terms of volume. The reason for this is essentially one of cost," Dicks writes.

"The 40G systems market has been held back because the price of systems has not been brought down to a point at which it is cost-competitive with 10G optical transport -- the cost of 10G systems and components, while not falling very fast, is still coming down."

So, the game is to prevent 100 Gbit/s from falling into that trap, by getting standards in place before mass deployment starts. On the Ethernet side, the Institute of Electrical and Electronics Engineers Inc. (IEEE) 802.3ba standard is expected to be completed this year, covering client-side connections for 40-Gbit/s and 100-Gbit/s Ethernet.

Perhaps more importantly, the Optical Internetworking Forum (OIF) framework for 100 Gbit/s is pointing optical component vendors in the same direction, toward DP-QPSK modulation with coherent receivers.

The report points out that standards typically aren't needed for line-side optics, because there's usually no need for interoperability there. But it seems clear that the industry wants a standardized approach, in order to keep costs down.

The result could be accelerated adoption of 100 Gbit/s. In an Infonetics Research Inc. study published in December, more than 40 percent of the carriers surveyed said they would be willing to deploy 100-Gbit/s links once the price is only twice that of 40 Gbit/s. That's earlier than would be predicted by the conventional wisdom, which says customers will buy a fourfold increase in speed at a price 2.5 times the previous generation.

Even so, 40-Gbit/s interfaces can't be written off. Dicks's report quotes vendors saying they don't expect 100-Gbit/s deployments to hit substantial volumes for at least two years. (See Is 100G Transport Closing the Window on 40G?)

The report covers the 40-Gbit/s and 100-Gbit/s strategies of 13 vendors, describing which parts of the network each is going after and how far they've gotten in deploying or developing their products: ADVA Optical Networking , Alcatel-Lucent (NYSE: ALU), Ciena Corp. (NYSE: CIEN), Cisco Systems Inc. (Nasdaq: CSCO), ECI Telecom Ltd. , Ekinops SA , Ericsson AB (Nasdaq: ERIC), Huawei Technologies Co. Ltd. , Infinera Corp. (Nasdaq: INFN), NEC Corp. (Tokyo: 6701), Nokia Networks , Nortel Networks Ltd. , and Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA).

— Craig Matsumoto, West Coast Editor, Light Reading

online123 12/5/2012 | 4:41:22 PM
re: 100G Standards Aim for Lower Costs

Could the fishy (too much hype for the result) CRS-3 announcement by CSCO  be prompted about 100Gbps fears?  If telcos are planning 100G now and CSCO lacks even a story, how can they be taken seriously as a core router vendor?  Separately, how much "routing" can be done at 100Gbps anyway?  At that speed, it seems that switching is the only thing possible.

Pete Baldwin 12/5/2012 | 4:41:21 PM
re: 100G Standards Aim for Lower Costs

Well, the big carriers already knew Cisco's 100G story (and probably know its multi-100G story and possibly even its 400G story).  I'm sure Cisco tries hard to not leave them in the dark.


It's true Cisco came out late, but Cisco claims it doesn't launch products without having a customer to name.  (Now, that doesn't mean they weren't just plain late to begin with, and we don't know how much arm-twisting AT&T might have had to endure before being willing to go public with the info.)


It appears that the big core-router vendors will all have 100G within the same time window, I think.  I think we'll have to look to other factors to decide winners and losers. (In a sense, that means there was EVEN MORE "too much hype" around the announcement.)


As for how much routing you can do at 100G... Good question. Processers certainly have gotten to the point of doing 50Gbit/s routing, and Cisco (and ALU, and possibly others) claims it's pushed that to 100G.  Does that stand up under prolonged real-world use?  Dunno.  I'd love to talk with someone who's taken a close look.

HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE