CyOptics - experts in chip design and automated packaging
February 24, 2006		Email this page

Interview with Ed Coringrato, President and CEO, and Stefan Rochus, Director of Marketing and Business Development.

Introduction

CyOptics, based in Breinigsville, Pennsylvania, was founded in 1999. The company has raised a total of $106 million in financing, most recently closing a $27 million series F round in October 2005.

CyOptics designs, develops and markets a range of optical chips and components for integration into access, metro and long haul communications systems. The company also provides contract design, fabrication and packaging services.

Overview

CyOptics is focused on the production and sale of Indium Phosphide (InP) chips and packaged components, primarily 14-pin butterfly and 7-pin GPO devices. Mr. Coringrato said that CyOptics is now turning its attention to the manufacture of small form factor TOSA/ROSA (Transmitter/Receiver Optical SubAssembly) devices for integration into transceivers. However, he emphasised that the company does not market transceiver, transponder or systems-level products.

Mr. Coringrato commented that CyOptics is an independent supplier to subsystem vendors and systems OEMs, adding,

"There is presently a great deal of discussion within the industry regarding the long-term viability of supplying discrete components to OEM companies, however, CyOptics is finding ample opportunities in this business. As an example, the company is currently tendering for new design wins with systems vendors seeking to use its components in their line cards".

CyOptics was founded by Uzi Koren, formerly of Bell Labs, who is now Chief Technical Officer at CyOptics. The company was established to develop 40 Gbit/s modulator and P-I-N receiver products, the market for which has not so far materialised.

Mr. Coringrato stated that CyOptics is proud of its 'tool box' - the building blocks and sub-components that can be employed in the development of new products, noting,

"The company possesses a unique pool of knowledge and expertise with respect to chip and component design and packaging techniques, deriving not only from CyOptics directly but also from twenty-plus years of research and development work conducted at Lucent and Bell Labs".

Strategic focus

Describing CyOptics' approach in more detail, Mr. Coringrato explained the two core areas of activity:

1. The design and manufacture of optical engines and chips supporting data rates from 155 Mbit/s up to 40 Gbit/s.

2. The provision of design and production services for optoelectronic components, including custom chip design, foundry work and packaging.

Regarding the production services side of the business, Mr. Coringrato observed that CyOptics is able to leverage its facilities and expertise to address the needs of both new start-ups and established vendors. He noted that many start-ups opt for an 'asset light', outsourced approach to component sourcing, while some established vendors simply do not have access to certain capabilities in-house,

"CyOptics' uniqueness in the industry derives from its expertise in chip design and production and its automated packaging technology - delivering unmatched precision placement and fibre alignment capabilities. This allows the company to differentiate its products via attributes such as wavelength stability. As the industry shifts to the use of more integrated components and higher data rates, these skills are becoming more sought after".

The company has chosen to restrict its activities to certain areas of the market but there are no technological issues preventing the development of, for example, transceivers. Operations acquired from TriQuint and CENiX included transceiver and transponder products, though CyOptics elected not to develop these.

On the issue of manufacturing transceiver products, Mr. Coringrato remarked that the sector is extremely challenging, with numerous players competing for business and yet more companies entering the market. For CyOptics this is seen to represent a valuable opportunity, as each of these transceiver suppliers requires components of the type CyOptics produces.

Commenting on the potential pitfalls of diversification and vertical integration, Mr. Coringrato said one factor that can affect the success of a small or medium-sized company is the need to stretch research and development resources across a wide range of product lines, spanning perhaps chips, components, subsystems, transceivers and line cards. CyOptics sees value in focusing its efforts on one segment of the market,

"CyOptics has a record number of products coming to market now, this is only possible because the company focuses exclusively on development of chips and components. Not only can the company offer a greater range of products, but also its components contain advanced next-generation chips that deliver valuable differentiation in terms of performance".

Mr. Rochus remarked that, while four to five years ago system vendors and OEMs may have preferred that component suppliers offer a broad product portfolio, including module products such as optical transceivers, they now generally consider product performance - along with cost - to be the most important attributes when sourcing components.

Acquisitions

Acquisitions have formed a core element in the evolution of CyOptics. In February 2003, CENiX, a developer of 10 Gbit/s transponders, was acquired, while in May 2005, the company acquired the optical components business of TriQuint - originally that of Lucent/Bell Labs - including a fab facility.

Although the CENiX transponder business as such was not of interest, Mr. Coringrato explained that the ability to provide packaging services using a highly automated process - and thus design and manufacture components - was. He added that CyOptics already at that time operated an InP fab in Israel to which the chip design and packaging capabilities of CENiX were added. The Israeli based fab facility that existed prior to the acquisition of TriQuint's Optoelectronics Division was surplus to CyOptics' requirements, and was subsequently sold to Vishay Intertechnology.

Packaging operations were later consolidated at facilities in Breinigsville, Pennsylvania and Matamoros in Mexico.

The foundry business

The foundry business is regarded as another diversification play for CyOptics. The capability is seen to give the company a unique position in the industry via the provision of specialised services to customers. Mr. Coringrato said foundry projects could prove process technology drivers for the company as well as straightforward sources of revenue.

The CyOptics fab plant is currently operating at around only thirty percent capacity, which is viewed as both offering an opportunity to expand the business and leverage existing infrastructure.

Mr. Rochus observed that while there are numerous contract manufacturers capable of providing packaging services, most concentrate on To-can manufacture and very few are able to deliver planar packaging services as CyOptics does. Mr. Rochus also noted that CyOptics is able to set up a production line quickly in response to customer demand,

"With regard to InP foundry services, it is important that a customer is provided with a turn-key solution, spanning epitaxy, wafer and chip fabrication. In general, companies in this business are able to deliver only one or two of these steps".

Offshore manufacturing

CyOptics' Mexican manufacturing facility was inherited with the acquisition of Lucent assets, having been established by that company in 1995. Initially, CyOptics used the site only for low-cost assembly work, but has subsequently relocated its automation platform to the facility.

To the observation that companies relocating manufacturing operations overseas today generally choose China, Mr. Coringrato commented that such a move can give a company a significant one-time gain in terms of labour costs, but that longer term the challenge is production yield, particularly with components for high-end applications,

"Beyond the initial benefit in terms of labour costs the issue becomes product quality and reliability - 'cost of quality' is a crucial factor for a company such as CyOptics. There is little to be gained from shifting production to a low cost location if the facility cannot then deliver the necessary quality. For CyOptics, the Mexico facility is the ideal compromise, as labour costs are low and the site is able to deliver the high quality necessary using the automation platform. For CyOptics' customers, product quality is a prerequisite".

Mr Rochus added that the Mexico operation is not as low-cost as China, but is probably on a par with certain Eastern European countries and Russia. He reiterated that production yield is of greater importance to CyOptics than straightforward cost of labour, adding that while the IP differentiation of a product is held in the chip, the majority of the cost lies in the packaging. The CyOptics automation platform enables the company to deliver high yields as well as high quality.

With regard to quality, Mr. Coringrato pointed out that CyOptics is expanding its addressable market to sectors beyond telecommunications, notably defence, where performance and quality are everything and cost is secondary.

Customers, competitors and collaborators

On the question of CyOptics potentially being viewed by customers as a potential competitor as well as a supplier, Mr. Coringrato said this is rarely a problem since the company structures each customer relationship according to the particular needs of that customer to avoid this conflict,

"In this context, concerns most often arise where competitors are customers of CyOptics - the issue being whether one company might discover what CyOptics is doing for another. CyOptics is able to demonstrate that it maintains customer confidentiality by firewalling activities carried out on behalf of each client. In this regard, the fact that staff are often dedicated to a specific customer project further reduces potential for information 'leakage'".

CyOptics cites the 'three Cs' when defining the market - customers, competitors and collaborators. However, many business relationships blur the boundaries between these definitions, especially as CyOptics works with and competes with companies ranging from start-ups to established players.

According to Mr. Rochus, other companies, such as TriQuint with its successful GaAs foundry business model, at times compete with customers, so this issue is by no means unique to CyOptics.

The system market is now mature and comparatively predictable, Mr. Coringrato observed, as opposed to the evolving but growing transceiver and transponder market. He explained that CyOptics still generates the bulk of its revenue from sales to system vendors, but market growth now derives from addressing transceiver suppliers serving the line card market,

"There are synergies to be realised from addressing both the system vendor and transceiver markets, for example there are opportunities to place next generation TOSA/ROSA devices directly onto line cards. CyOptics is currently involved in projects to accomplish this".

It was noted that the established system houses have a long history of working with CyOptics, both in its own right and via its acquisitions of the assets of TriQuint, Agere and Lucent. The company is increasingly engaged in discussions regarding the design-in of components to new systems, and in addition has accompanied transceiver suppliers on visits to their system vendor customers. The company is equally comfortable addressing either level of customer, Mr. Coringrato remarked.

The 40 Gbit/s market

On the topic of the much-vaunted 40 Gbit/s market, Mr. Coringrato said he personally expects the technology to present a volume opportunity for components suppliers in 2007, adding that current technology and processes may offer cost advantages over those available five or six years ago. CyOptics is presently assessing different approaches to 40 Gbit/s systems, including array technology.

The relationship between cost and the adoption of 40 Gbit/s technology is also an area CyOptics is investigating. Mr. Rochus observed that existing 40 Gbit/s interfaces are typically designed for short-haul applications - over a maximum distance of 2 kilometres - for linking ADM and router equipment. This is believed to present a barrier to wider deployment of the technology, as costs escalate dramatically when extending reach to 10, 40 or 100 kilometres.

Mr. Rochus explained that the cost of deployment of 40 Gbit/s equipment rises as the link distance increases due to the need for Raman amplifiers and dispersion compensation technology - both of which add substantially to the overall cost of the network.

To address this issue, CyOptics is applying CWDM technology, combining four 10 Gbit/s links, for the transport of TDM traffic in metro area networks. This approach offers the advantage that a 10 Gbit/s line rate is used and therefore only issues relating to 10 Gbit/s transmission, using proven and economically viable technology, need be considered. CyOptics believes a range of at least 40 kilometres, and possibly up to 80 kilometres, can be achieved using this approach.

CyOptics is presently engaged in a project with Cray to develop 80 and 160 Gbit/s technology and the company anticipates being able to apply elements of this development work to 40 Gbit/s products.

Regarding 40 Gbit/s technology developed in 1999 to 2000, when CyOptics started out, Mr. Rochus remarked that this work is still valid, and also now includes the results of research conducted by Lucent and Agere. He cited, as an example, an electro-absorptive modulator and receiver for 40 Gbit/s applications that could be marketed if demand existed, adding,

"Customers today would be interested in 40 Gbit/s technology for deployment over 40 or 80 kilometre distances if this could be achieved at lower cost than using four individual 10 Gbit/s links. One obstacle to this, as far as cost is concerned, is the need for Raman amplification and polarization mode dispersion compensation. However, the fundamental 40 Gbit/s technology developed five or six years ago, at the chip and packaging level, is still potentially relevant to the market".

Optical Keyhole conducts interviews on the basis of readership interest only. They are not paid for by the participating companies, nor is there any swap for newsletter subscriptions or advertising.

This article is the copyright of Optical Keyhole. It may be freely distributed by any means in an unaltered form.