Katherine Butler
Predictive Models for Transceiver Performance
Whitepaper Overview
The diverse assortment of components and technologies that work together within the optical transceiver make the process of constructing a roadmap challenging. Often, this diversity is cited as an insurmountable barrier that makes any picture of the future necessarily murky. In this white paper, we demonstrate a methodology for using physical models for roadmapping.
The starting point for this analysis has been a clear definition of the physical parameters that are important; in this paper we take signal performance (for example bit rate) and power dissipation (Watts consumed) to be the important metrics. A model for the entire transceiver is developed that establishes which devices are responsible for the majority of the power dissipation for a given performance. As we shall show, more than 80% of the power in a high performance 10 Gbps transceiver is dissipated in just three components (multiplexers, thermoelectric coolers, and driver amplifier). The influence of associated devices on the power is also modeled; the laser and TE cooler as well as the modulator and modulator driver.
Once we have sifted through the various components to identify those critical to signal performance and power, we consider the alternative materials and technologies for each component. Low-level physical models are used to explore these alternatives. Where possible, data from existing research and established roadmaps is incorporated into these models.
The eventual output of these models is a clear picture of transceiver performance, component materials, and manufacturing processes for the future. These physical models are eventually integrated with the technical cost models, economic models, system dynamics models and cognitive models developed in the CTR collaboration. Read More...
Thesis