https://www.linkedin.com/feed/update/urn:li:activity:7163574533185093632/
An additional reason for Marvell Technology’s relatively early announcement of its 800ZR/ZR+ solutions in the summer (please see: https://lnkd.in/ePeR2it6), would be as a preemptive action to get potential competitors of the modules to think twice before entering the space. The avoidance of as many players as happened at 400ZR (totaling at least around the mid-teens), resulting in tremendous price pressures, would be in the interest of all optical component/system suppliers. There might be an expectation of a lower number of participants anyway given that some of them may pause to see if 1600ZR is feasible within a satisfactory period (please see: https://lnkd.in/e8JTmpb3). Additionally, although Meta had unexpectedly come up with plans to be a big buyer of 800ZR (please see: https://lnkd.in/e8V9N5RF), fibeReality has learned that there will also be an especially healthy amount of purchasing involving 800ZR+ pluggables – that makes sense, with a re-architecture of its optical network. The ZR+ would inherently offer a higher margin to Marvell, as the hypersclalers generally have not tended to desire that gear. While Amazon was an exception with 400ZR+ -- and would be anticipated to be interested in 800ZR+ -- fibeReality is unaware of any firm plans for that device by the customer at this time.
Turning again to Ciena and 400ZR (please see: https://lnkd.in/exCFY4bQ), there could be a simple reason for its lack of substantial penetration -- a decision by the vendor to avoid treating the device as a loss leader. With 800ZR, there may be a greater chance of resulting in an acceptable margin for the firm.
