Symmetricom’s Sync Business Not Necessarily a Sure Thing

Implications: 1.Symmetricom has just about a monopoly on synchronization gear in the U.S. 2.While alternative distribution methods might open the door to other competitors, the supplier’s domineering position might remain. 3.However, there is a possibility that the RBOCs might ultimately move away from Symmetricom as their sync vendor.

Analysis: Even as new suppliers move to synchronous Ethernet or IEEE 1588, most of them are not providing the precise oscillators – and that is where Symmetricom would still come into play -- in providing the clocks. However, the RBOCs believe that Symmetricom has abused its corner on the market by arbitrarily discontinuing products to force them to buy their new systems at much higher prices – in order to recover its R&D costs. There have also been issues around service and overall attitude on the part of the supplier. The manufacturer has developed a reputation for not listening to its customers and strictly just telling the RBOCs its intentions. In addition, Symmetricom has apparently still not met all of the Telcordia GR requirements for its latest product – despite manufacture discontinuing the previous generation back in 2002.

Qwest considers FEI’s product to be really good and it is running it through lab testing – and there is the potential that it could standardize with this vendor for distribution – getting a better solution at equal or less cost. In fact, the Qwest’s lab people have been working with FEI quite a bit in general. FEI has also helped itself by both adding more elements to its overall package as well as hiring an ex-sync expert from the old Pacific Telesis, who is of course really familiar with the standards and requirements of the RBOCs.

Oscilloquartz has also gotten more serious about penetrating the States in terms of getting it through the NEBS process for fire and RFI considerations. It has a really attractive product for small applications. Still, the vendor’s unfamiliarity with the processes in the U.S. has been a hindrance. There is one experienced individual out of Colorado Springs working with Oscilloquartz.

If Qwest decides not to stay with Symmetricom, the two other RBOCs might follow because Qwest is the only one of the three that still has people left with extensive knowledge of sync. Another problem for the manufacturer is that these types of products are looked upon like battery plant in the telco environment – they are critical – but as long as they are working – do not mess with them. They are not usually replaced until there are problems -- despite the fact that Symmetricom latest device is remote accessible. Suffice it to say, it would be hard to peg the company as a high growth opportunity.

Symmetricom might be really picking up on the potential trouble for its traditional business. It recently hired a person to make a personal visit to one of the RBOCs – an action that has not been taken there in quite a few years – and he received a huge earful of complaints.

At best, Symmetricom has to deal with the spurts that have occurred in demand for their sync equipment – as the RBOCs put off buying as long as possible. After being ignored for a while, the realization sets in that they have to purchase a bunch of the gear. In general, there will probably not be a mad rush to buy new systems – partially because the older versions tend to last very long in the network. For example, both Sprint have Qwest have sync equipment that was deployed in the early 1980s. And the FCC committee study in 2002 that showed that about 10 percent of the reportable outages were because of sync, is probably high now. One of the reasons for the decline has been a shift early in the decade from electrolytic capacitors on the power supplies to the use of ceramic.

Qwest in particular is not chomping at the bit to spend as much as $90 million to replace its network. The RBOC is in great shape in warehousing the old technology. It essentially has all it needs as it tried to anticipate not only repair, but also growth.

Concerning the future, Symmetricom has been trying to push in standards meetings the idea of putting a time stamp with Ethernet as opposed to a sequence number for reassembling packets. If it was to become successful at that, the industry is going to need to have some type of an NTP standard. Still, once X number of those is out there and people know how to access them to be able to establish their time, then the time can be kept internally – resulting in much fewer NTP servers.

In getting precise time on the stamps, one needs about a half second accuracy – so that one is on the right second. So what is really being talked about is thousands of a second – but that is still only 10^-3 – so, the same type of accuracy is not required. Quartz can handle it as opposed to rubidium or cesium – with some exceptions such as the special requirements of government.