• Technical Conference:  30 March – 03 April 2025
  • Exhibition: 01 – 03 April 2025
  • Moscone Center, San Francisco, California, USA

Will Linear Pluggable Optics (LPO) Have a Future Beyond 112G?

Sunday, 24 March, 16:00 – 18:30

Room 6F

Linear drive targets the elimination of DSP/retimers between host ASICs and optical engines for potential cost/power/latency benefits. Linear drive is well suited to co-packaged optics (CPO) or near package optics (NPO) applications that offer tight integration with short electrical links between the IC and optics. However, there has also been intense promotion, advocacy and compelling demonstrations of linear drive applied to traditional pluggable optics. Although the results for linear pluggable optics (LPO) are encouraging, challenges remain to be overcome to enable widespread adoption. Examples include signal integrity limits and specifications, interoperability, testability, and ecosystem establishment. This workshop will provide a forum for sharing opinions and insights from different segments of the ecosystem: end users, standardization bodies, and suppliers of ICs, transceivers and optical components, and systems. This workshop will address the following questions.

  • Which use case will LPOs be highly possible to land?
  • Will there be a broad window to adopt 100G/lane LPOs? Are 200G/lane LPOs feasible?
  • What are the limits imposed by signal integrity and what specifications are needed?
  • Will LPOs have a strong impact on deployment, commissioning and operation practices for end users?
  • Are AOCs a more promising form factor to implement linear drive?
  • How much power/cost/latency savings are expected for LPOs?
  • Will LPOs be ecosystem friendly?
  • Will interoperability between LPOs and traditional pluggables be required?
  • Will LPOs transfer design complexity from transceivers to hosts?
  • Which types of optical engines are best suited for implementing LPOs: VCSEL, SiPh, DML or EML?
  • Will it be feasible to push toward direct drive in pluggables by further eliminating in-module drivers?
  • Will new modulator materials (TFLN, BTO, etc.) help enable the adoption of LPOs and the evolution to 200G/lane?
  • Will LPOs be a step on the way to CPO or an alternative to CPO? How much know-how and standardization legacy from LPOs can be transferred to CPO?

Organizers

Jiangqiang Li, LightsAI, United States

Andreas Matiss, Corning, United States

Katharine Schmidtke, Eribel Systems, United States

Clint Schow, University of California, Santa Barbara, United States

Speakers

Andy Bechtolsheim, Arista, United States

Chris Cole, Quintessent Inc., United States

Ryan Latchman, MACOM, United States

Ashkan Seyedi, Nvidia, United States

Davide Tonietto, Huawei, Canada

Chongjin Xie, Alibaba, China

Qing Wang, Meta, United States

Xiang Zhou, Google, United States