PORTLAND, Ore.--Silicon photonic pulse compressors were demonstrated on CMOS chips recently by researchers at the University of California at San Diego (UCSD). To be used for optical time division multiplexing (OTDM) in future on-chip silicon photonics, the pulse compressor design used a dispersive optical grating to generate short, powerful communications pulses on a CMOS chip.
UCSD claimed its pulse compressor is the first CMOS implementation strong enough for OTDM.
Future CMOS chips will communicate over optical interconnects, using pulse compressors and other on-chip photonic components to aggregate individual data channels into high-speed optical signals. Besides speeding up the interconnect over copper wires, optical signals also reduce power consumption and run cooler than electrical signals. The compressed pulses aggregated by UCSD's device were seven-times shorter than the pulses in the original signals.
Scanning electron micrograph (SEM) of dispersive grating for generating short, pulses of light for optical interconnects. Credit: UC San Diego / Dawn Tan.
UCSD's approach used a integrated dispersive element designed by UCSD doctoral candidate Dawn Tan, and which is now a part of the CMOS photonics toolkit being built by UCSD professor Yeshaiahu Fainman. The technique transforms long pulses of light into spectrally broader, but shorter, pulses that can be multiplexed simultaneously over a single optical channel. The researchers predict that terabyte per second data rates will be achieved with their technique.
Dawn Tan, the Ph.D. candidate in the Department of Electrical and Computer Engineering at UCSD's Jacobs School of Engineering who led development of the pulse compressor. Image credit: UC San Diego / Daniel Kane.
Isn't a laser diode silicon based? like those used on CD players.
Though I too wonder how is the light generated or transmitted in to the CMOS chip?
Perhaps is all done through optical interfacing... optical fiber is used perhaps and the compressor receives such light and optically transforms it in to another set of frequencies but shorter in time duration, hence increasing the speed. Interesing article of what's happening in sunny San Diego.
Congrats to professor Yeshaiahu Fainman at UCSD, Dawn Tan and the research team...this is interesting step towards optical communication between CMOS chips...but the fundamental problem remains: how do you generate light on a CMOS chip? there are NO silicon based lasers to my knowledge...Kris
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.