Regardless of a news report of a very low purchase price for BTI Systems, and despite Juniper Networks’ contention about the acquisition “not [being] a new strategy,” the move is actually a watershed moment for the corporation, and it does not seem to reflect its historic reputation for investments in a deliberately thoughtful manner with regard to the long-term ramifications of such a purchase, as it is based more on short-term expediency to try to satisfy Wall Street investors now. There is a huge difference between “developing optical interfaces…on [its] routers” and delving totally into the optical systems space, which Juniper has previous avoided because of its lack of comfort level in getting too far beyond the IP piece. In addition, while the vendor’s assertion that the buyout will be quite targeted in “captur[ing] what [it] believe[s] are very significant market inflection points” sounds appealing, it ignores the economies gained by competitors in offering solutions with common parts across the board.
Also, although the BTI “acquisition is about technology and talent, [and] Juniper doesn’t anticipate it will have a significant material impact in 2016,” getting fully into DWDM will mean a good amount of investment in R&D. We believe that BTI placed more emphasis on marketing concepts, such as its cloud focus, than on practical engineering concerns, including in making minor modifications on its 10G technology to remain more competitive, including in potentially giving it more flexibility to lower its prices. Of greater concern is that in order to stay in the game for Data Center Interconnect (DCI), cash to produce higher capacity/lower density offerings will be necessary. Furthermore, Juniper now places itself in the position of being responsible for supporting BTI’s entire installed base of equipment.
It is a reasonable assumption that Cisco Systems will become one of the leading DCI providers, especially at the high end. Therefore, in the overall share rankings in the DCI market, BTI will likely be no greater than in sixth place. In fact, although it was a metro supplier at one time at one of the largest hyperscale customers, it has been reduced to only supplying amps and other small components there. Apparently, this loss of business at this user appears to be the major reason for the company abandoning its plans to go public.
A major problem may be Juniper’s lack of sufficient expertise in the optical transport business. If the speculation about an exceptionally disparaging price is true, incentives to retain executive talent at BTI may be problematic.
Of course, another matter that we have discussed in the past surrounds the uncertainy about the actual growth potential of the DCI space. While one research firm is touting over a four-fold increase over the next several years, it is difficult to imagine the foundation for such a claim. Actually, an experienced engineer in the optical engineering space commented on our last blog post on LinkedIn as follows: “At OFC [two] years ago, an analyst told me that if anyone was selling me numbers on the size of the data center interconnect market, they were lying. It’s impossible to get a true accounting….”
[written by Mark Lutkowitz]