Regular followers of fibeReality’s separate, quick-update blog were able to read our latest take on Infinera, as we stated: “We view the messaging [on its last webinar], as a clever means of slowly weaning the industry off the stress on PICs, by using language associated with these chips, as the company moves forward in new directions.” Given the fiasco associated with Cisco Systems’ acquisition of Acacia Communications, the requirement to move the attention on the company away from high-end optics in general becomes more pronounced, as it will not only need to wait for help from a different merchant DSP supplier for its Groove platform, but the odds of it creating an ICE6, and shipping it in volume, after giving up on ICE5, should probably be considered out of the realm of possibility. Thus, in the same way that the “Nortel” optical culture lives on at Ciena, the indication is that there will be a restoration of the “Tellabs” mindset at Infinera, which would be refreshing, and a step in the right direction, as it would gradually cease being the sole, optical system supplier advocating for its PIC technology, which has obvious limitations. Moreover, for those top-level networking executives at the incumbent ISPs, who have been around for at least a couple of decades, they will remember “Tellabs” fondly, as a supplier, which tended to have a solid reputation for being reliable and responsive. As always, the extensive operational ties resulting from the “Tellabs” gear in their installed base are undeniable. The fundamental problem is that it is not clear at all that there is sufficient technical and/or cost differentiation at the moment to construct a long-term, viable business case, even as Infinera expands into new areas, such as with open routers and with targeted automation tools.
In early 2008, the CEO of “Tellabs” stated: “[O]ur focus on data has been not to…be the fourth router supplier in an already crowded market, but it has been on finding niche applications, developing the appropriate software that enables these applications and selling the product on the basis of value, and thereby holding the margin.” Yet, this strategy did not save the company from eventual collapse, starting with AT&T deciding to give a good amount of this business to Cisco. Later, despite its investment in a new 9200 edge router for the mobile backhaul market, the product became unsustainable.
At least, Infinera acknowledges that we are in the very early days of an open ecosystem for routers, and so, we would argue that the extent of the desirability, especially with traditional operators, still needs to be determined, regardless of any early, PR announcements by such users. At least on the optical side, there is the well-known problem of being unable to ensure five-nines reliability with such an approach. Additionally, the fact that it is has shipped over 300,000 IP routers becomes less relevant, if there, is once again, an inadequate means to adequately distinguish the DRX, other than its openness.
Concerning automation, and putting the hyperscale data center operators aside, we agree with Infinera that despite all of the attention over the years in favor of IP migration and SDN controllers, the incumbent service providers are still dealing with extensive, antiquated operations support systems — and we would add plenty of PSTN/TDM elements as well. Especially, as it relates to OSS, it has been recognized by infrastructure architects for a long time that there cannot be a wholesale rip-and-replace, and the introduction of various forms of practical automation tools will be a very slow, incremental process. So, while it is a quite positive that Verizon appears to be attracted to the Transcend solution, there are both big suppliers, such as Ciena, as well as small ones, including Sedona Systems, which are also moving in the same way as Infinera.
Furthermore, we do not necessarily agree with the significant use of pizza boxes by any of the large, traditional ISPs in the States. In fairness, though, it can never be ruled out that AT&T may be so inclined to go cheap and dirty. Conversely, we wonder if there are still potential networking considerations in lacking a full chassis – not to mention, the potential problems, for sure, in employing a 2RU rack to mount 1RU hardware.
All in all, while such a radical change for Infinera this late in the game probably helps to further guarantee future survival, inevitable, sustainable growth is a whole other matter — nor after all of this time, can a true “Tellabs II” absolutely be put together, after losing so much of the original management. Nonetheless, we believe that Robert Shore, Senior VP of Marketing, obliquely made somewhat of a convincing case on the webinar that a quasi-“Tellabs”-centric company could at least hypothetically become a reality.
Still, as far as we are concerned, “Tellabs” never recovered from the shenanigans that took place behind the scenes resulting from the merger with Ciena failing to happen, and especially later, when the former failed to come up with an answer to the CoreDirector acquired by the latter. Anything resembling a “Tellabs II” needs to significantly adjust for that major hole from the past. There is also no doubt in our minds that if the late Rob Pullen been appointed CEO much earlier — in 2004, history would have been much different for that corporation.
As always, fibeReality does not recommend any securities, and this writer does not invest in any companies being analyzed by us.
[written by Mark Lutkowitz]