During the summer of 2017, to the best of our knowledge, there were major technical problems at Apple with Lumentum’s six-inch VCSEL wafers (including with the epis supplied to the component vendor). Startlingly, after a short period of time, Lumentum Holdings apparently produced them in high volume with good yields. Despite the firm’s remarks to the contrary, pulling off this extraordinary feat without full control of the development/manufacturing process within the company itself made it that much greater an achievement. Even more astounding, the vendor did not milk this monumental accomplishment for all it was worth on its last quarterly earnings call, especially after Finisar made it clear that it was pretty much a novice in addressing the challenge beyond four-inch VCSEL wafers. Actually, the size of these devices was not brought up during that conference call at Lumentum, and there was continuing discussion about offering a separate solution as well, edge-emitting lasers. Regardless of our opinion that Lumentum at times lacks the frankness of other large optical component firms (such as its insistence that it is still interested in the 100G CWDM4 market), and in spite of our current view, which is contrary to the present industry rhetoric, that four-inch VCSEL wafers cannot be ruled out in playing an even more important role with 3D sensors moving forward, the supplier has to be given the benefit of the doubt that everything it shipped was the six-inch variety.
Despite the fragile nature of the bigger offering, we never precluded the possibility. Broadcom, which is far out on the learning curve when it comes to VCSEL bare die seems quite confident that it could produce the larger type of wafer, while being fully aware of the technological obstacles in the way of the present players. Fortunately, for the current competitors, we firmly believe right now that the chip powerhouse does not see it as worth its while to enter the 3D-sensor space.
As we have discussed in the past, at least on paper, there is obviously a greater bang for the buck with using six-inch wafers. However, from a practicality perspective, it is certainly much easier to work with the smaller gear.
While Lumentum has a large lead at the moment in this consumer market, we maintain our stance that there are numerous signs that the confidence level of Apple is increasingly shifting toward Finisar. While we realize there were political implications, we understand it was Apple’s idea for Finisar to purchase the new fab in Texas.
It has also been extremely rare over history for Apple to make joint announcements with its component suppliers. The iPhone vendor not only did so with Finisar, but the anticipated orders at the time exceeded those of Lumentum.
We also picked up some intelligence last summer that Finisar initially held firm that it could not make money in shipping four-inch wafers to Apple. Given the early challenges at Lumentum, we presume Apple was forced to pay a premium to Finisar. Finisar’s adamant stance at the time was probably another factor in helping to earn Apple’s respect.
The good news for Finisar is that we believe this experience in supplying four-inch wafers may have resulted in the realization of a win-win scenario with Apple. In fact, we would not be surprised to see at a minimum, some smaller devices being produced for Apple out of that new plant, despite it being dedicated for six-inch wafers, especially with Finisar’s new CEO being in a good position from an expertise standpoint to reinforce a compelling case.
Getting back to the topic of candor, Lumentum referred to the current drop-off in demand as “seasonality.” Such a characterization seems at least a little disingenuous to us in that it likely has more has to do with Apple, and perhaps other potential customers, taking some time to reassess as to what extent they want to bet on demand for facial recognition as well as other applications enabled by 3D-sensing.
To follow us on our separate, daily company blog, which is exclusively on fibeReality’s LinkedIn page, please click here.
[written by Mark Lutkowitz]