Power and Performance of POET’s DML Solution
January 12th, 2022
In December 2020, POET announced an industry first. Operating at speeds of 25 gigabits per second (GPS), POET’s family of four directly modulated lasers (DML) of different wavelengths are the first known commercial 25G DFB-type DMLs to utilize a flip-chip process to passively align and bond to electronic and optical circuitry, while maintaining optimal performance.
The flip-chip assembly technique and the unique POET Optical Interposer platform enables a true single-chip, fully integrated Optical Engine to be produced at wafer-scale. The result is the lowest-cost, smallest-size 100G CWDM4 Optical Engine with a form factor of 9mm x 6mm. It includes banks of four lasers, four monitor photodiodes, four high-speed photodiodes, a multiplexer, demultiplexer, taps for power monitoring, and features that support a self-aligned fiber-attach unit (FAU).
POET’s process reduces the cost of building photonics devices by 25% to 40% compared to conventional approaches.
The only DML integration platform in the world wasn’t created in a vacuum either. It was achieved because leading enterprises in the optoelectronics industry tasked the POET team with finding a way to deliver greater performance without requiring the retrofitting of infrastructure or an increase in costs. The DML platform has quickly proven to be adaptable at 100G and 200G CWDM4, as well as for LR4 devices. The company states that the next big step in its evolution is to adapt the DML platform to 400G.
“We’re taking a two-step approach to 400G,” CEO Dr. Suresh Venkatesan recently said in a video Q&A with investors on the Reddit forum. “We’ve got our one step on the CW laser with modulator, but we’re now starting to look at the DML integration. We’re saying, 'Hey, we’ve got this platform. Nobody else does. It’s working really, really well at 100G and 200G. How about we see if we can just extend this to 400G.’ If we do that then we get a quick slingshot into that market.”
Venkatesan believes the DML solution can extend to 800G with the potential of reaching speeds of 1.6 terabytes and 3.2 terabytes. Incorporating qualified lasers from highly regarded manufacturing partners accelerates the market readiness of the POET Optical Engines.
The technology’s flip-chip assembly of photonics and electronics with passive alignment enables automated, chip-scale packaging. Being able to flip-chip lasers allows POET to assemble Optical Engines at wafer-scale, which is the single most important driver of reduced cost.
Given the estimated total available market (TAM) for 100G transceivers of approximately $2.5 billion, POET believes that Optical Engines based on the Optical Interposer can reach annual revenue of over $100 million within this single market segment in the 2025-26 time frame. It can meet the anticipated demand thanks largely to its joint venture company, Super Photonics Xiamen (SPX), formed in cooperation with Sanan IC, whose parent, Sanan Optoelectronics is one of the world’s leading compound semiconductor manufacturing companies.
The flip-chip assembly process allows POET to readily incorporate lasers and other active devices into derivative Optical Engine configurations, supporting data communications applications such as 200G CWDM4, 100G CWDM6, and 100G LR4, telecom applications such as 5G, as well as other applications that could benefit from the small size and low cost of the Optical Interposer platform.
To achieve the flip-chipping of lasers, POET had to demonstrate that it could simultaneously optimize the RF (radio frequency) performance of the flip-chipped DML on the interposer while preserving a low RIN (relative intensity noise) measurement both before and after assembly.
(Note: RF performance relates to the quality and power of the electrical signals, while RIN is a measure of stability of the lasers that are assembled on the Interposer.)
The DML platform and the elegance of the Optical Interposer design allows for multiple integrated Optical Engines to be placed on one module. That feature achieves greater functionality while reducing the bill of materials by up to 80% over several existing alternatives. Such innovation and efficiency are among the key reasons why POET is working with blue-chip and medium-sized companies in Asia, North America and Europe. Volume production is on track to begin in the third quarter of 2022 and customer engagements continue to grow and deepen as the DML platform demonstrates its accomplishments. [Learn more]
< Back to Blog