Since the beginning of the arrival of the 40G data rate, there has been a consistent mantra in the industry of its impending decline and ultimate demise. Large market research firms have supported this point of view as they received the vast majority of their business from the dominant transport equipment vendors, which concentrated their product development as well as public relations efforts on 100G and beyond. A principal beneficiary of the continuing good health of 40G in the data communications space is Applied Optoelectronics (AOI), which remains the dominant 4x10G transceiver supplier. Given the company’s strong position in this sector, including a relatively high level of vertical integration, price pressures at that speed have been minimalized, and helps to create a buffer against its gross margins perhaps getting any further down than the low-to mid 30s in the foreseeable future.
Even AOI has up until recently been among the chorus characterizing 40G in the following ways: “certainly in the tail end of the product life cycle,” the “transition of the mega-datacenter customers…to 100G,” and the expectation of “a gradual decline.” However, it turns out that AOI’s revenue in 40G will “[probably be] higher in 2017 compared to 2016.”
It is hardly out of the question that as the overwhelming bulk of the data center operators in the world, which are operating at no higher than 10G, will at least in some cases be attracted to the lower cost 40G before moving to 100G. Consequently, the 40G business may not drop significantly for an indefinite period of time.
Fully realizing that the Street tends to focus narrowly on the cutting edge, sexier story of 100G and higher, AOI on its last quarterly conference got defensive about the crossover point from 40G to 100G. “[W]e don’t really talk about that….I can only speak to our numbers…[W]e’re seeing very strong results both for 100 gig and 40 gig. 40 gig actually started with a higher base. So the growth rate there is less. 100 gig is growing very strongly. But I can’t really comment on the overall industry outlook. I don’t have that data.”
Interestingly, on that last earnings call, AOI broke out the percentage of its CWDM and PSM data center revenues, “as a point of reference that we do not expect to provide on a go-forward basis.” In addition, the firm stated “for AOI, CWDM products generally have higher gross margins than PSM due to our advanced design and manufacturing capabilities for these products.”
Yet, it seems very clear that the main reason for only disaggregating the revenue only one time as well as stressing the higher margins on the CWDM is the artificial lowering of prices of the PSM4 (100G), which has made any kind of reasonable margin generation probably impossible. Also, even if AOI has a cost advantage with CWDM4, there will continue to be substantial competitive price pressures on this type of device as well.
Along with the cost of development of new products at longer reach and at higher data rates, the stabilizing effect of its 40G portfolio of solutions will be even more critical to AOI.
[written by Mark Lutkowitz]