Our internet-enabled world has created a global race for ever-faster, cheaper and more efficient communications. A lead driver in this race is David Plant of McGill University. A pioneer in lightwave research, he focuses on creating high-performing optical technologies with record-breaking data transfer rates.
For example, last fall the Professor of Electrical and Computer Engineering demonstrated how members of his laboratory were able to achieve speeds of 224 Gbits/sec using a single laser and modulator. “It was the first time this rate has been achieved in this way,” says Plant. It is a significant accomplishment because the single-laser setup offers a more scalable solution for meeting the increasing demand for faster, inexpensive, short-reach optics.
Since 2007 Plant has held the NSERC/Bell Industrial Research Chair in Ultra-High Bit Rate Optical Transport and Access Networks, which has placed him in the forefront of developments dedicated to creating hardware for the Internet to grow and flourish. His research is increasingly important as networks struggle to keep up with the soaring demands of power- and bandwidth-hungry applications. In 2013 he earned a prestigious Killam Research Fellowship, which has supported his ability to tackle this challenge in new and creative ways.
The key to his research achievements, he says, has been a long-standing technology that was once regarded as an exotic laboratory tool: silicon photonics.
“It’s a platform that allows you to build complex, high-performance photonic circuits that are useful for enabling short-reach fiber optic interconnects,” he explains. “We’ve applied what are proving to be some original signal-processing concepts to an emerging technology to produce transmission capacity on a single wavelength that is setting high watermarks.”
For Plant, this approach has been a game-changer that has seen a great deal of technical capability, previously only available to large multinational players, and put serious innovative potential within reach of a university-based facility like his.
“Silicon photonics is now becoming a foundry-accessible platform that doesn’t require you to have your own specialized fabrication facilities,” he observes, although he does acknowledge that his laboratory has been very well outfitted with the support of the Canada Foundation for Innovation. “You combine the best-in-class of infrastructure with some really bright students and fireworks happen.”
Acknowledging that progress in the speed of short-reach interconnects (for example, from 500 meters to 10 km) has long been a solution in search of a problem, Plant says the right problem has finally come along with the rise of data centres. These dense clusters of high-performance computing arrays have emerged as the crucial nodes for major network traffic driven by on-line service giants. These facilities can easily cover hectares of ground, so optimizing their operation means enabling computers to exchange information as quickly as possible.
“Data has to go from one side of the warehouse to the other over distances of up to 10 kilometres, something that is not possible using electronic alternatives,” says Plant. “You want to do that on optical fibers.”
He credits CMC Microsystems with sustaining an interest in the technology that has allowed his team to create industrial applications that have drawn private-sector partners into the research. “CMC is amazing,” he insists. “They never let go of silicon photonics.”
With the support of CMC, he adds, his laboratory has acquired Ericsson as a sponsor. “They’re interested in linking with us because we’re doing things that they foresee as being relevant for their future.”
Among those things has been dispersion engineering, which takes advantage of frequency-dependent signal delays to enhance the quality and quantity of long-distance data transfers, and the design of new types of optoelectronic VLSI circuit systems that have streamlined the way data is handled within or between devices. This background was complemented by CMC’s support for his research on silicon photonics, which he regards as just another step toward building tomorrow’s Internet.
“We have been able to do hardware work that five years ago—in the absence of CMC—simply wouldn’t have happened,” he concludes.