When it comes to 3D-sensors, both Lumentum Holdings and Finisar are being gutsy in striving to make a tremendous amount of penetration, not only away from their traditional market strongholds, but also in a new product space, VCSEL dies, (arguably even a somewhat brand, new experience for Lumentum), which has technology challenges not meant for the fainthearted. Those individuals following our other blog have been able to keep up with some of the potential minefields in being a chip vendor to Apple’s iPhone including 1) only a minimum of two vendors needed for Apple to lower prices to shockingly low levels (and other smartphone competitors will expect their charges to be in the same ballpark); 2) the effects of iPhone pricing strategy itself; and 3) the increasing willingness for Apple to bring chip production in-house, and even leasing manufacturing equipment out, if a particular vendor drops out, because it could not justify the margins. Another factor potentially limiting demand at Apple is that close to two-thirds of iPhones formerly purchased continue to be used. Nevertheless, the technical hurdles, especially with 6-inch wafers, used by Lumentum give us the most pause.
Fundamentally, it is hard for us to totally rule out disqualification because of power and heat issues with the use of all VCSEL dies in general in the iPhone. However, when it comes to 6-inch wafers, we understand that at least at one time, II-VI had overheating issues, which may help to explain why it has been late to the party.
While Lumentum expresses great confidence about its six-inch wafers, which has to be taken seriously, it would be fantastic if the external production by Win Semi has not had to overcome some tech matters. The gentleman in the comments section of our blog article: “Finisar: 3D-Sensing Opportunity in Context,” expresses it very well concerning the “two risks in running VCSEL on six inch wafers. First, there is an inherent challenge in the epi layer. Second, there is a challenge in handling the thin and brittle wafers.”
Over history, the problem for pioneers of all types is the high risk in being shot. While Finisar is supposedly deciding whether to move beyond its 4-inch wafers, and adopt 6-inch devices itself, manufactured externally, internally, or both, it may be buying time to see how well Lumentum is able to pull off such a high-volume shipment.
If there does turn out to be some difficulties, Finisar can sort of come to the rescue with its wafer design. In addition, Apple would probably be inclined to like giving the vendor additional business because of Lumentum’s poor judgement in advertising its gross margins involving a customer, which is always preoccupied with getting that last penny of cost out of its smartphone.
While we think it is unlikely that Lumentum ends up with a warehouse full of VCSEL bare dies, Finisar, which appears to have a much smaller level of commitment at the moment, has a reduced level of risk for either late changes in features by Apple on a particular iPhone model, or just the sheer abandonment of a certain functionality. For example, there has been speculation that facial recognition may not be included in the next release, which would apparently eliminate one of the big needs for these 3D-sensors.
[written by Mark Lutkowitz]