Combatting the Age of Big Data
Telecommunications companies – those that allow us to talk on the phone, communicate over the Internet and watch cable television – used to operate under the notion that there was an infinite amount of fiber bandwidth available to transmit these signals. Then we moved into the Y2K era.
“There was a big explosion of data around the year 2000,” said Larry A. Coldren, the Fred Kavli Professor of Optoelectronics and Sensors at the University of California, Santa Barbara. “Computers were also getting faster and faster at this time and the demand for bandwidth was rising quickly.”
Coldren and his team had started looking at photonic integrated circuits (PICs), devices that allow signals to travel on optical waves on semiconductor chips, back in the 1980s and discovered that they could viably be produced much like analogous electronic integrated circuits (ICs) that generally use electrical wires for transferring data.
Last month, he spoke about his research during Boston University’s Electrical & Computer Engineering Distinguished Lecture Series. He suggested that PICs could be the key component in the future of telecommunications.
Just a couple of decades ago, wavelength-division multiplexing (WDM) was introduced to meet the demand for more fiber bandwidth. This method allowed a number of signals to be simultaneously transferred on a single optical fiber. However, at the terminals where the WDM channels must be either combined or separated, the optical and electronic equipment became more and more complex as the channel count and signal speed increased. That’s where Coldren’s research comes into play.
“PICs have the potential of improved performance, reliability and cost while also reducing the size, weight and power of the equipment,” said Coldren.
PICs for various applications have been made using indium phosphide, silica on silicon, polymer technologies, and silicon photonics. Electronic ICs, however, usually use silicon as a dominant ingredient. Coldren’s team currently focuses on a monolithic indium phosphide integration platform.
“Ultimately, we may find that the best results will come from a hybrid solution using more than one of these materials,” said Coldren.
Today, PICs are widely deployed commercially and outperform many discrete device approaches, but Coldren is optimistic that they can work even better in the future and hopefully result in more environmentally friendly supercomputers and data centers.
“Our efforts have always been focused on making PICs very efficient and very fast,” said Coldren. “Now we need to look at how they can be used to create more green data centers.”
Assistant Professor Jonathan Klamkin (ECE), who introduced Coldren at the lecture, previously had an opportunity to study with Coldren while earning his Ph.D. at UC Santa Barbara.
“I benefitted immensely from his guidance and even use his books in my class here,” Klamkin said. “It’s a pleasure having him on our campus.”
Prior to teaching, Coldren worked at Bell Labs, where he studied surface-acoustic-wave signal processing devices and tunable coupled-cavity lasers. He continued his work at UC Santa Barbara, where he has developed more widely-tunable DBR lasers and efficient, high-speed vertical-cavity-surface-emitting lasers (VCSELs) in addition to his PIC research.
Coldren is a member of the National Academy of Engineering and a Fellow of the Institute of Electrical and Electronics Engineers (IEEE), the Optical Society (OSA) and the Institution of Electrical Engineers (IEE).
Coldren’s talk was the third in the three-part Fall 2013 Distinguished Lecture Series. The seminars will resume in Spring 2014.
-Rachel Harrington (email@example.com)