Stewart Wills

23 March 2017—The American Institute for Manufacturing Integrated Photonics (AIM Photonics) yesterday announced public release of the 2016 Integrated Photonic Systems Roadmap (IPSR), a massive, 400-page study that seeks to capture market drivers and technology and investment needs for the U.S. integrated-photonics industry over the next 15 years. AIM Photonics developed the study in collaboration with the MIT Microphotonics Center and the International Electronics Manufacturing Initiative (iNEMI).

The roadmap—resting on input from some 700 photonics-industry professionals from 16 countries over a period of 18 months—sees a particularly large market impact for the much-heralded Internet of Things (IoT). It also frames some specific requirements, especially in training, tools and standardization, to meet the technical requirements likely to play out in the next decade and a half.

Dizzying demand growth

At the heart of the roadmap is the recognition that the appetite for data, in both the United States and globally, appears inexhaustible. The report cites well-known market estimates by Cisco that annual global data traffic, pegged at more than a zettabyte (a billion terabytes) in 2016, could double by 2020, marking out a compound annual growth rate of 22 percent.

To meet that kind of demand growth over the longer, 15-year timeline covered by the IPSR, the report’s authors estimate that the components making up the global network face some sobering requirements. Those needs include 10,000-fold improvements in power efficiency and cost per function, and a million-fold increase in the number of network ports—all, the roadmap document dryly notes, “at no increase in total cost.”

Strategic concerns

Some of those improvements, the roadmap authors note, can come from changes in basic processor technology and architecture, such as the replacement of common blade servers with emerging microservers and further pushes to Moore’s Law. But a significant amount, particularly to achieve power reductions, will need to occur in the optical domain via interconnects and optical switching, according to the report. And the study envisions an environment undergoing a broad “paradigm shift” driven by cloud-connected devices and the IoT. That shift, it suggests, will raise new challenges in business models, security, product innovation and more, and will increasingly require ongoing development of “disruptive” technologies to handle.

Within that environment, the IPSR stresses the need to strengthen cooperation among researchers, industry, and other stakeholders toward a coordinated vision, to smooth the road ahead. The development of standards—cutting across areas including manufacturing, packaging, and even workforce training—to ensure interoperability and market potential constitutes an important underpinning of such a vision as articulated in the roadmap.

More broadly, the 2016 IPSR identifies four large strategic concerns for the coming environment for integrated photonics. These include a lack of government and other structures to support manufacturers undertaking the risks implicit in developing disruptive technologies; the persistent competition from copper-based interconnects, and striking the right market balance between copper and fiber; the difficulty of developing a cooperation framework for industry amid intellectual-property concerns; and the pervasive need for standardized, commoditized components to meet cost targets.

PEGs and TWGs

The business end of the roadmap consists of reports from six “product emulator groups” (PEGs) and nine “technology working groups” (TWGs). Each PEG report projects technology needs for a specific product or business sector; those PEG analyses are then used as input for the TWG reports, which constitute mini-roadmaps for specific technology development initiatives such as monolithic integration, packaging, design and training.

The Internet of Things looms large in the work of the PEGs; indeed, one of the emulator groups focuses specifically on the IoT. But in addition to the report from that IoT PEG (notable for an unusually high density of exclamation marks), the market dynamics envisioned in the analyses from three of the other product groups—data centers, automotive products, and photonic sensors—all rest on a large helping of demand attributable to IoT-driven paradigm shifts.

Roadmap recommendations

Based on the input from the PEGs, the TWGs offer a number of recommendations in the roadmap for prioritizing efforts in integrated photonics over the next 15 years. Among the more intriguing include:

• the need to better develop college- and graduate-level training in design of integrated-photonic technologies;
• the parallel need for better integrated design tools;
• basic improvements in manufacturing technology to support miniaturization, cost-effective assembly and other imperatives;
• materials development to support monolithic integration and improved performance and reliability;
• and, as an underlying theme, standards development to tie the whole bundle together.

The roadmap is a considerably richer document than this brief account may suggest. The full study, and the executive summary, can be downloaded at http://photonicsmanufacturing.org/2016-integrated-photonic-systems-roadmap.