+44 (0)24 7671 8970
More publications     •     Advertise with us     •     Contact us
 
Technical Insight

More Growth in Communications

Convergence of Optics and Electronics Last year, we chose InP Microelectronics as one of our "Things to Watch", and there was certainly a great deal of activity in this field. TRW announced that construction of its new 4-inch InP production facility was completevolume shipments of InP chips are expected in the second half of 2001. Furthermore, both Vitesse and RF Micro Devices revealed that they plan to use InP ICs for their next generation products, focusing on 40 Gb/s fiber-optic components and wireless handset power amplifiers respectively. One of the hottest topics in the industry at present is the convergence of optical components and communication ICs. There are several possible levels of integration, ranging from the co-packaging of electrical and optical components in modules right through to monolithic integration of different types of devices on a single wafer. And while CMOS, SiGe, GaAs and InP are all contenders for the ICs needed in such modules, only InP is a viable choice for the emitters and detectors at telecom wavelengths. Therefore, despite its relative immaturity, InP s importance in our industry is going to continue to grow. This will be especially true when companies begin looking for integration platforms for 40 Gb/s components. Convergence: A New Hot Topic While integration has been a popular research topic for 15 years or more, it has suddenly become a topic of intense interest in the commercial realm. That was demonstrated in December when the investment firm JP Morgan organized a one-day conference on the topic of the convergence between optoelectronics and communications ICs that drew over 500 attendees! Several topics such as next generation packaging of optics and ICs, contract manufacturing of optical products and the integration of optics and electronics on a single substrate were discussed. One of the first areas of convergence is packaging of optical components and ICs in the same box to create transponders that perform send and receive functions for any optical port. These types of product have already been introduced by suppliers such as Lucent and JDS Uniphase. Charlie Whilhoit, an analyst at JP Morgan, believes that commercial 40 Gb/s optical modules will emerge towards the end of 2001 or early in 2002, while the real momentum for these products will develop in the second half of 2002. Lou Tomasetta, President and CEO of Vitesse, told the attendees at the JP Morgan conference that "at 40 Gb/s, the module is going to be the component". The system manufacturers simply are not going to be willing to purchase 1520 different components from 510 different suppliers and then take on the task of assembling it all. Instead, Tomasetta said, they want to buy "a box with an optical fiber coming out the end." What are the drivers behind this trend? The first is cost. In the current system there are too many layers of overheadthere is the component company, the module assembly company, one or two other contract manufacturers, and, ultimately, the system company itself. In addition, more highly integrated modules would produce savings in size and power, and can also deliver improvement in performance. The Roadmap for Integration What is the roadmap for integration? Tomasetta says the place to start is detectors and amplifiers integrated together. Next is electroabsorption modulators with high voltage drivers, a classic example of something that can be done in InP but can t be done if you are using a lithium niobate modulator or are making your drivers from GaAs or SiGe. "After that you can begin to integrate the modulator, the driver and a wave guide amplifier. Then you can do the detectors, the amplifiers and the physical layer devices, the MUXes and DEMUXes. Eventually, you can integrate a transceiver and sooner or later integrate even a transponder, with the laser probably being the most difficult component to integrate," said Tomasetta. He said the goal is to reduce the cost of a module from its current level of $5,000$10,000 to less than $1,000. "I know a lot of people say that is not possible, but with high volume integration on four, six or eight inch wafers, you would be amazed at what kind of cost reductions you can get once you invest in doing the integration." Whilhoit predicts that a number of strategic partnerships or mergers will arise, cutting across the boundaries between traditional electronic and optical component manufacturers. Increasing investment is also likely to be targeted towards companies that focus on packaging, modules and integrated components, rather than those producing stand-alone devices. All of this may sound as though there is a juggernaut gathering steam in the communications industry. Nevertheless, don t be surprised if you hear some contrary opinions. "Typically the most prevalent technical argument against integration is that the best technology is needed for each component and customers are not willing to trade off any performance," explains Tomasetta. But he notes that this is generally an over-simplification. There may be some very-high end applications where this argument is truefor example, a very long single reach cable with no amplifiers in the middle. However, for the high-volume, short-distance applications, the system companies will be happy to trade performance for the many benefits of greater integration.
×
Search the news archive

To close this popup you can press escape or click the close icon.
×
Logo
×
Register - Step 1

You may choose to subscribe to the Compound Semiconductor Magazine, the Compound Semiconductor Newsletter, or both. You may also request additional information if required, before submitting your application.


Please subscribe me to:

 

You chose the industry type of "Other"

Please enter the industry that you work in:
Please enter the industry that you work in: