Andrew Schmitt’s presentation, entitled “Mapping the Future of Optical Deployments,” was one of several featured during this years ECOC Market Focus sessions in Gothenburg, Sweden September 18-20th.
The presentation highlights Cignal AI’s perspective on how different sectors of the market increasingly demand more specialized solutions. It also highlights the weaknesses of the much-discussed “Optical Whitebox” as well as the strengths and weaknesses of hardware disaggregation. Finally, we warn those in the industry to beware of The Cult of Open.
Subscription not required – complete the form with a corporate email address to download the presentation.
"*" indicates required fields
Takeaways from the presentation include:
- The optical equipment market is divided; there are several submarkets with very different technical requirements. There is no one-size-fits-all solution, and what works well for Microsoft will not work well for Verizon. Therefore, extrapolating the needs of one customer market onto other customers is a mistake.
- The divisions of this market now create conflicting requests to equipment and component suppliers who cannot undertake every desire of all customers. They are looking to concentrate their R&D firepower where they see the highest return; in the past 18 months, this has been the cloud & colo operators as well as more conventional operators looking to adopt the same DevOps approach to optical networking.
- Optical spending growth from cloud and colo operators is tilting the playing field (and the conversation) toward their requirements. While still a relatively small share of global spending, cloud and colo is the growth market, and vendor investment follows growth.
————-
- The white box approach was successful in hyper-scale environments for servers and Ethernet switches because it solved critical needs not met by the commercial equipment market. In the WDM transport space, it isn’t clear what needs are not already being solved by the eight different equipment vendors building compact modular equipment specifically for this application.
- These compact modular systems are now the fastest growing portion of the market and are seeing adoption across multiple operator types.
- Incumbent operators (Telco) still account for the majority of spending, and it is vital to understand how different their requirements are from cloud and colo operators (Webco). These unique needs of Telcos are what led to the development of the Packet-OTN equipment market.
—————-
- Deployment models are not discrete choices but are instead a spectrum of approaches ranging from sourcing from a single vendor (vertical integration) to utilizing open architectures (dis-aggregated hardware), all the way to designing one’s hardware from scratch (white box).
- There is a vocal contingent within the industry advocating for white box hardware without acknowledging the complexity of most customers’ requirements. We call this “The Cult of Open.”
- There are pros and cons to every deployment approach, and ultimately each vendor selects a deployment model best suited to its unique culture and requirements.
- Equipment vendors will continue working towards providing access to their hardware via open software interfaces. Simultaneously, vendors must find proprietary approaches that differentiate their equipment to address the customers who wish to build vertically integrated networks from handpicked vendors.
- The more rapidly that operator service and business models change, the more likely operators are to pursue open optical networks.