Toward the end of last year, Cadence announced that it is working with Lumerical Solutions and PhoeniX Software to create a Virtuoso custom design platform-based design flow for electronic and photonic ICs. Using this integrated electronic/photonic design automation (EPDA) environment, designers can take advantage of a schematic or layout-driven flow for both photonic and electronic circuits, photonic component parameter extraction and model generation, photonic circuit simulation, and photonic mask layout implementation for monolithic as well as hybrid 3D-IC photonic circuits. One of the principal minds behind Cadence’s push into photonics is Gilles Lamant, a distinguished engineer at the company. I recently sat down with Gilles to talk about why now is the prime time for photonics to move into mainstream electronic design. First, what drew you to engineering as a career? In the ‘80s back in France, where I’m from, I was an Apple II Europlus user and had been an Apple fan for many years—that got me into the field. Contrary to a lot of people, I chose a path in EDA. I have a master’s degree in electrical engineering, with focuses on device modeling and signal processing. During my final master’s project, I worked at Cadence as a co-op student. Since serving my military duty, I’ve been with Cadence ever since, with stints in Japan (8+ years) and Russia (two years). What sparked your interest in photonics? An important part of my work is to look at what’s next. For example, about seven years ago, when FinFETs were sort of an experiment, I began examining how the Virtuoso platform could support FinFET designs. About three years ago, I started looking at photonics. We try to anticipate the challenges that our customers will bring to us…so we’re looking at spintronics, graphene sheets, and other fun things. Why is now the time for the mainstreaming of photonics? Photonics is not an emerging technology—that is a myth I would like to turn off. What is not yet mature in this area is the EDA environment. And what’s really driving photonics to the mainstream is the huge bandwidth demands of applications like communications. Telecom companies with datacenter servers are now looking at as much as terabtyes of data to connect all of their backplanes in the datacenters. This is a huge amount of data to connect via fiber optics, and photonics is how the processing and memory are hooked onto these backplanes. Besides the backend of the communications network, there are a slew of potential other application areas that can benefit from photonic ICs. Sensing and sensor applications come to mind, for instance. What are some key techniques designers can apply to overcome the challenges of developing integrated photonic ICs? We’re trying to change the current way of doing things, to bring into photonics a more mainstream design flow where there’s a documented methodology. The right documented methodology today ( the EPDA environment ) is close to an electrical methodology, so designers who must move into photonics will be in a familiar environment. We want to make the transition as easy as possible for designers. With the standard one-off tools that are available, the learning curve is huge. Design teams know the Virtuoso platform. With Virtuoso, they don’t need to learn a whole collection of new tools for simulation, layout, and so forth. Can you share any insights into any emerging developments in the area of photonic IC? We’re examining various developments. One is the true co-simulation of photonics, thermal, and electronics, which is challenging because the time scales are very different for each. We’re also looking at using some better technology for 3D modeling for more accurate model extraction. We still want to have a basic framework for a flow that is familiar to designers. So there are some interesting challenges as well as potential solutions on the horizon. Christine Young![]()
![]()
↧