Given that the lion’s share of the discussion in telecom forums is about vaporware, it is not surprising that a lot people in the industry would probably be amazed to hear that R&D spending on 10G devices is still happening. While the really compelling drama within suppliers is occurring with mature or even declining market situations, including on 1G and on VCSELs, in which substantial revenue is on the line, the obsession on analyzing the accounts of the various futuristic, sexier solutions can sound repetitive, and at the end of the day, they have a tendency to be quite boring. In contrast, the struggle for 10G transport equipment vendors to survive is a captivating narrative involving a data rate, which will remain a mainstay in public and enterprise networks for a very long time -- in which there is still an overabundance of participants (partially because financing going back to the bubble has allowed them to hang around), and prices have dropped to levels that were never close to early expectations. (Amazingly, in 2013, it was not hard to find a really long-reach 10G interface with optics for under $5,000.)
The vastly underappreciated growth in 1Gs in the future will eventually require aggregation into 10Gs. Demand for 10G tunable gear, including the form factor, SFP+, is expected to grow noticeably. In addition, the use of the 10-gig speed will probably increase substantially in the vast majority of data centers in the foreseeable future as well as represent the primary rate for switches in that space.
With such big incentives and despite all of the attention given to 100G, transport system providers have kept their eyes on the continual progression of 10G motivated by attaining the lowest cost and the highest density – including the proper pluggables, the precise engineering rules, and the best architecture. There has also been relatively recent development on some really innovative and very low-cost 10G solutions.
Unfortunately, as part of the 10G survival process, system suppliers are forced to be distracted because of the pressures from their customers, the press, and industry analysts to develop credible stories for solutions that are likely to remain almost exclusively vaporware in the market for an indefinite period of time. Particularly in the public networking business, they include 400G and SDN capabilities.
We have stated in a recent blog post that it will be at least 10 years or more before there is widespread network deployment at the 400G rate. Given the historic pace of change, a safer bet is likely closer to a couple of decades out. Moreover, the drive to get a satisfactory ROI on their 100G investment will be the paramount concern for optical system vendors before development money substantially shifts to the next data rate.
We have also addressed our reservations about SDN, especially concerning the non-enterprise sector, as well as bandwidth on demand. At least with the large incumbent service providers (somewhat related to movement to 400G), the psychological resistance to change is a key consideration. Furthermore, SDN is to a large extent, a relabeling of where the industry has been supposedly working towards in the management and the interoperation of networks.
[written by Mark Lutkowitz]