XXX XXXXX MATTHEW PEACH E-O INTERFACE
sophisticated digital signal processors to be able to interpret and decode and do all of the signal processing at the electrical level and for that we are using the Acacia’s DSP products.” SMALLER FORM FACTORS LeMaitre agrees that the well- documented trend of optics and electronics elements becoming more compact and power efficient is driving the revision of modulation formats into new dimensions. “We have to be able to do this in conjunction with key developments in electronics such as the transition to 7nm digital signal processing in order to be able to increase performance while allowing it to be packed into smaller and smaller form factors,” he said. “Our job at Oclaro is to try to deliver the best photonics engines for these and then find the right partners that can give us the electronics sitting next to it. We do not develop the electronics so we have to rely on the industry and the DSP suppliers. Our part is based on using InP but there are other ways to do it; for example, there are other groups working on silicon photonics.” OPTICAL INTERNETWORKING Optical Connections interviewed the OIF (Optical Internetworking Forum) about its involvement in the development of electro-optical interface technologies. The OIF, which this year celebrates 20 years of activities, promotes the development and deployment of interoperable systems through international implementation agreements for optical networking products and component technologies. 100Gbps lambdas and therefore the industry will benefit from a 100Gbps electrical interface definition that enables industry interoperability while providing the simplest implementation of an optical transceiver function. Our CEI-112G projects are in the process of defining electrical channel definitions for die-to-die, host-to-module, chip- to-chip and backplane interoperable implementations.” The Forum commented, “Next generation optics will operate at
Some of the OIF’s current development projects that are addressing the industry’s optical and interface needs include the following: The High Baud Bandwidth Coherent Driver Modulator project prioritises RF performance over size by defining a small form factor component integrating a high performance, polarisation multiplexed quadrature modulator plus the RF drive functions for the high baud-rate and low modem implementation penalty segment of the coherent market. The Integrated Coherent Transmit- Receive Optical Sub Assembly project combines previous OIF efforts to define The primary area of interest is around the implementation of coherent technology integrated photonic components for both transmit and receive functions into a single package suitable for the next generation of high-density optical modules and onboard optics. The 400ZR project is an OIF optical project designed to specify a digital coherent module interface targeted at short reach DWDM links that enables interoperable, cost effective 400Gb/s implementations. Instead of specifying new technologies, the project provides a lighter version of a long haul, 400Gb coherent link. REVOLUTIONARY RESEARCH Focusing on the importance of the electro-optical interface, a research team comprising members from City University of Hong Kong, Harvard YVES LEMAITRE CHIEF STRATEGY OFFICER, OCLARO
University, Ma, US, and Nokia Bell Labs have fabricated an on-chip lithium niobate modulator that is smaller, more efficient and enables faster data transmission than previously available. The research project, entitled “Integrated lithium niobate electro- optic modulators operating at CMOS-compatible voltages,” was first announced in Nature journal in autumn 2018. The technology is said by its developers to be “set to revolutionise the industry.” This electro-optic modulator measures just 10 to 20mm long and its surface area is about 100 times smaller than traditional devices. It is also highly efficient, having higher data transmission speed with data bandwidth tripling from 35 GHz to 100 GHz, but with less energy consumption and ultra-low optical losses. Existing and commonly used lithium niobate modulators require a relatively high drive voltage of 3V to 5V, which is significantly higher than 1V, the voltage typically provided by CMOS circuitry. This requires a complementary electrical amplifier, which makes the whole device bulky, expensive and energy hungry. Project leader Dr. Wang Cheng, assistant professor in the Department of Electronic Engineering at CityU (Hong Kong), commented, “In the future, we will be able to put the CMOS right next to the modulator, so they can be more integrated, with less power consumption. The electrical amplifier will no longer be needed.” Thanks to the advanced nano fabrication approaches developed by the team, this modulator can be tiny in size while transmitting data at rates up to 210 Gbps, with about 10 times lower optical losses than existing modulators. This revolutionary invention is now on its way to commercialisation. Dr. Wang believes that those who look for modulators with the best performance to transmit data over long distances will be among the first to get in touch with this infrastructure for photonics. CONCLUSION Whether the physics of construction materials will ultimately limit E-O modulator performance to a matter of terabits per second, other innovations are likely to enable greater speeds, particularly the innovation of parallelism. Oclaro’s Yves LeMaitre concluded, “Probably the next step in my view is that a lot of the developments will be about parallel solutions until there is the next serial technology breakthrough. Somebody smarter than me will come up with the next serial step. 1 Tbps seems possible in terms of existing electro- optical technologies but beyond that, at the moment the only way seems to be to go parallel so I think we will continue with progress via a series of serial and parallel developments.”
Oclaro is focused on EO developments based on an indium phosphide platform
www.opticalconnectionsnews.com
15
ISSUE 15 | Q4 2018
Made with FlippingBook - Online catalogs