x

Optical Crossconnect Surge Seen

News Analysis Light Reading 8/28/2001

Market research firm KMI Corp. says worldwide revenues from optical crossconnect systems will quadruple this year, reaching $6.3 billion by 2006
(see KMI Has Good News, Bad News).

Over the next year alone, sales of optical crossconnects will increase more than 100 percent, KMI predicts.

As sales increase, the per-port price of optical crossconnect equipment will fall nearly an order of magnitude, the firm says. By 2006, the average price per port will be $1,100, versus today's $10,000.

According to the report, demand for optical crossconnects is being spurred by the rise and complexity of data traffic, which carriers are finding expensive to handle in the same fashion as voice traffic.

With traditional Sonet gear, circuits are "nailed up" or permanently fixed to carry high volumes of traffic through carrier networks. This "one size fits all" arrangement is great for voice. But data is supported by gradations of bandwidth, designed for a range of applications and priced accordingly. With Sonet equipment, carriers must manually set up individual optical channels in order to accommodate these differences. The result is nothing short of a nightmare for many providers.

Enter the optical crossconnect, which KMI says is helping carriers save up to 70 percent on network costs, compared with traditional Sonet equipment. (For primers on optical switching and crossconnects, see Optical Switching and Optical Crossconnects.)

Today’s optical crossconnects come from a range of vendors, including:

BrightLink Networks Inc.

Calient Networks Inc.

Cisco Systems Inc.

Lucent Technologies Inc.

Nortel Networks Corp.

Sycamore Networks Inc.

Products from these and other companies are based on a range of components and technologies, but most of them perform OEO (optical-electrical-optical) conversions inside the box. By 2006, KMI predicts, most crossconnects will be based on all-optical designs, allowing for greater speed and capacity per unit.

— Mary Jander, Senior Editor, Light Reading
http://www.lightreading.com

COMMENTS

Newest First | Oldest First | Threaded View

Add Comment

jr 12/4/2012 | 7:52:28 PM

re: Optical Crossconnect Surge Seen Cinta Networks offers a product with integrated DWDM transport and OXC. Are you familiar with their product? Do you know of anyone else with a similar offering? When do you think the market will be ready to embrace an integrated solution?

Reply | Post Message | MESSAGES LIST | START A BOARD

Light2001 12/4/2012 | 7:53:53 PM

re: Optical Crossconnect Surge Seen All the shipped cross connect (or optical core siwtch, using RHK's definition) are electrical fabric based. The expensive part is really the linecard, which contains data-rate specific optical transceivers. For OC48, the cost now is ~ $10-13K, but for OC192, the cost will be at least $25K. This doesn't count the actual switching fabric and common equipment. To drive the cost of linecard down to $1000 is just simply impossible.

So my guess is that the cost KMI is refer to is really "bare bone" 3D MEM switches, which did not count any linecards, but only switching guts, plus the control cards. The cost of these switches depends on the size of the switch, and the per port cost for 1000x1000 is infinity,, since no body shipping products. Good luch to see the cost go down to $1000 any time soon.

The over all requirement for the caiier may still be right, in the sense that they do need these OXC or OSCs to manage networks and derive revenues by providing fast bandwidth provisioning and efficient network protection and restoration. It's both a cost saver, and money maker.

To really derive cost effective OXC though, one need to look at the products that integrated DWDM transport with the OXC as single network element, so maximum of one OEO is needed. One example being that using tunable laser is take care of wavelength switching, and use photonic switch for fiber switching. Before large NxN photonic switches are available, small 1xM switches can be used.

Reply | Post Message | MESSAGES LIST | START A BOARD

Mark Sebastyn 12/4/2012 | 7:53:57 PM

re: Optical Crossconnect Surge Seen This is a big load of BS.

"With traditional Sonet gear, circuits are "nailed up" or permanently fixed to carry high volumes of traffic through carrier networks. This "one size fits all" arrangement is great for voice. But data is supported by gradations of bandwidth, designed for a range of applications and priced accordingly. With Sonet equipment, carriers must manually set up individual optical channels in order to accommodate these differences. The result is nothing short of a nightmare for many providers."

Hellllllooooo KMI? Ever heard of virtual concatenation? Optical Crossconnectects nail up (giant inhale) wavelengths and SONET nails up 'individual optical channels' (giant exhale).

Can someone (especially someone from KMI) explain the difference? I'd rather nail up STS-1's than wavelengths.

WHEN WILL THE HYPE END?

Reply | Post Message | MESSAGES LIST | START A BOARD

wdog 12/4/2012 | 7:53:58 PM

re: Optical Crossconnect Surge Seen $10,000/port going to $1000/port. What kind of port?

The forecast numbers are really silly.

But the idea is right. Using circuit switches (optical cross-connects) for traffic engineering of bandwidth from the metro across the core is the most realistic approach for carriers. With or without GMPLS

Reply | Post Message | MESSAGES LIST | START A BOARD

Pseudopersonality 12/4/2012 | 7:54:00 PM

re: Optical Crossconnect Surge Seen What is the definition of a "port"?

Is that a bi-directional port i.e. two physical connectors on the switch or a single physical port?

What is the reference criteria for a port? Is it a n OC-3, OC-12, OC-48, OC-192, or other?

Is this cost per port with common equipment included or just the added circuit pack?

Are there any studies to justify the $6Bil 2006 or is that just a swag?

Thanks for your attention

PP

Reply | Post Message | MESSAGES LIST | START A BOARD

[email protected] 12/4/2012 | 7:54:00 PM

re: Optical Crossconnect Surge Seen This seems rather interesting given where we are in the market. Ciena (who has the bulk of the "optical switch" market has now warned us that they are unsure of their market for the future, the long haul carriers (who will be buying the bulk of this stuff) are in some tough financial shape and we have seen two significant sub-systems people Trellis and OMM either abandon ship or quit the high end business. What's more is that the price/port seems to be low so one would assume that we are looking at super high port counts.

The conclusions seem like the same old conventional wisdom that we have heard for selling the all Optical story. I also like the chart in this story....quite a ramp.

Could double as a ski jump.

Reply | Post Message | MESSAGES LIST | START A BOARD