BLOG

Tibit: Last “Telecom Valley” Optic Hurrah

August, 2019

 

Over the years, nobody can deny that some of the most prominent “Petaluma Telecom Valley” entrepreneurs/engineers were courageous in taking on the huge, technological challenges, in the inherently hardest sector to generate meaningful and sustainable profitability, from the perspective of both optical infrastructure suppliers and service providers: the access portion of the network. It has always been difficult because much of that “last mile” involves residential traffic across a range of incomes, but regardless, there has increasingly been an overall propensity on the part of these subscribers, in general, to expect superior service, without a significant rise in the price. There has also been the classic struggle to gain easy, direct entry into these homes. Nevertheless, the Petaluma approach was to take the new technology to the next stage of complexity, while still aiming to substantially lower the total cost of the devices. So, there was Advanced Fibre Communications, which focused its development initially on digital loop carrier equipment targeted to non-tier 1 urban areas, actually over copper, as well as Teknovus, on the enterprise side, which offered EPON on a chip. It is the team from the latter, which makes up the technical foundation for Tibit Communications, the most ambitious undertaking yet in the Valley, with a pluggable PON, and as fibeReality briefly discussed recently, we believe the justification for extensive use of its solution is questionable. The net effect of establishing the precedent of “data center economics” in a traditional operator cost structure will, paradoxically, result in rapidly diminishing returns on investments for such daring achievements in the optical realm, particularly at the highest end of the food chain. Therefore, we anticipate that Tibit will be the final such example to have even the possibility of becoming viable commercially in the access space, as the ability to attract sufficient interest upfront, from VCs, should become very tough.

For the hyperscale data center operators have not even begun to be finished in their quest to reach the absolute bottom on pricing of optical transceivers (TXs), even after the downfall of the data communications ecosystem. A prime example is Facebook, under the guise of innovation, is employing the 100G CWDM4-OCP (the cooperation of Arista Networks in this effort also helps explains its transition in rhetoric to the lower data rates) as a ruse to tighten the margins of component vendors even more, as the same optics found in the standard, two-kilometer CWDM4 specification will wind up being used in both solutions. Thus, in effect, the operator is obtaining 2km capability at zero cost to itself.

In looking at Tibit’s OLT a couple of years ago, the engineering issues discovered with the MAC in each of its pluggables, which consumes a lot of power, as well as the thermal budget, are still both in play. In fact, the OLTs are about twice the size of the length of an SFP, with quite big heat sinks. Transitioning to higher “layers of switching and traffic management” hardly mitigates these disadvantages.

It is also instructive to look at China, which is well-known for its widespread deployment of PONs, and the operators in that country desire combo OLT ports, allowing them to accommodate the various generations of the passive technology from a single TX. They find there is a greater efficiency in having a MAC on a line card catering to multiple PON ports, rather than making it increasingly more difficult by including a MAC in each TX.

Additionally, in comparing an early press release from Tibit to its most recent one, the company is definitely late in making an ONT available. Moreover, although NG-PON2 has been on its roadmap, we are unaware of any evidence of the firm moving in that direction, other than becoming a member of the Broadband Forum.

For our fast-growing, totally separate, quick update blog, which is exclusively on fibeReality’s LinkedIn page, please follow us here.

As always, fibeReality does not recommend any securities, and this writer does not invest in any companies being analyzed by us.

[written by Mark Lutkowitz]

 

Share Article

Add Comment

 

Back

 

Read Comments