Unfortunately, this writer will be unable to be in Rome, as I will be attending a much happier, personal event later in the week. fibeReality thought it might be a good idea to propose some hard-hitting questions that our loyal readers might consider asking suppliers at the annual hype-fest, just to shake things up, if nothing else. Here are 24 suggestions:
1. To any supplier, associated with COBO, is it not time to move on to co-packaged optics, which although would not be easy to pull off, it would be a more practical means of accomplishing the chip-to-chip concept, if it ever does happen in the foreseeable future?
2. For Macom, is it not a little disingenuous to say that lasers in your PICs are integrated, since the former cannot be accomplished with silicon photonics, and is the company adequately able to differentiate itself from products offered in a similar 200G/400G announcement by Semtech?
3. For Inphi, was your worst mistake getting so involved with Microsoft, which led to the creation of ColorZ, only to later face a lack of differentiation prospects with the standardized ZR approach at 400G?
4. To Infinera, will you be able to leave what (in fibeReality’s opinion) has been from the very beginning, a weird culture of superiority, which has not been justified, resulting in at best, becoming a second-rate supplier, and there will not be another opportunity to save yourself with another acquisition?
5. To Lightwave Logic, is it not a red flag to be claiming high performance in a small, compact package, while comparing your solution to Lithium Niobate instead of InP?
6. For Huber+Suhner, how much more patience will you have on slow, high-cost, all-optical switches being used in a significant way?
7. For Oclaro, are you likely to become the guinea pig for Lumentum, as a means of experimenting in new businesses, such as for wireless support?
8. To Finisar, was bringing in a key technology expert from Nokia, a part of a greater effort to enter the fronthaul space, to run the coherent effort, or for another reason?
9. To Lumentum, will the leadership be prepared to put aside what we believe to be hubris, and fully embrace the talent and technical prowess from Oclaro, while totally appreciating Apple’s push for an RFIC foundry model for VCSELs, which will lead to tremendous margin deterioration?
10. For Mellanox Technologies, have you really given up your silicon photonics solution?
11. For Corning, with your hands full with producing singlemode fiber, and therefore, multimode fiber not fitting well into your business model, will manufacturing of all of the latter be farmed out to Fujikura?
12. To Luxtera, given your relationship with TSMC, why has there been barely a peep (see our comment below on the link) about co-packaged optics, and is it not the case that discrete optics would do the job just fine?
13. To Ericsson, are you able to give an affirmative answer on meeting Verizon’s demands for overseeing the requirements for OSS with Calix?
14. For Juniper, how much of a hit are you taking on margins in being under Ericsson’s umbrella for active infrastructure for Verizon’s 5G wireless network?
15. For Cisco Systems, if the service provider space has more downs than ups moving forward, what will be the potential impact on the optics business, when even bigger products, like routers, would undoubtedly be considered for sharp cuts in operating expenses?
16. To ADVA Optical Networking, will there still be adequate room for a viable niche player with sufficient margin generation in the long term, even with the full collapse of the optical ecosystem caused by the hyperscale players?
17. To NeoPhotonics, is it time to consider a drastic move, such as paying a premium to an executive, like Greg Dougherty, assuming he is not put on the board of Lumentum?
18. For Intel, did you not leave the optics space several years ago?
19. For Google, even if you provide the funds for an OSFP form factor to get to 800G, to be realized in the very long term, does it make economic sense from a volume perspective, if you are the only major user?
20. To Facebook, have you come to the realization that if you truly want co-packaged optics to happen, you will have to make large investments yourself?
21. To Nokia, after re-architecting your router with a 2.4T ASIC for the MPU running at 56G SerDes, in order to more efficiently accommodate Broadcom’s Tomahawk 3 chipset (getting to the full capacity of every port of 400-gig), is there not an issue with density in running the board on the old-chassis down at 25-gig SerDes to get to that high data rate?
22. To Ciena, in moving up the stack with packet capabilities, is it not an admission that new optical development cannot be fully paid for by the business itself?
23. For Sumitomo, what is the real confidence level in making multicore fiber happen, especially with a point-to-point connection using silicon photonics chips, which would solve the problem of having to break out the fiber cores?
24. For Credo, how can you honestly believe that 400G will happen on the client side next year?
As always, fibeReality does not recommend any securities, and this writer does not invest in any companies being analyzed by us.
To follow us on our totally separate, quick update company blog, which is exclusively on fibeReality’s LinkedIn page, please click here.
[written by Mark Lutkowitz]
SHARE THIS