Greed stifles DOCSIS chipset Market - In-Stat/Insights

Electronic News, Feb 25, 2002 by Eric Mantion

Many different kinds of cable modems (CMs) are available, but at the core of all of them are chipsets based on the data over cable systems interface specifications (DOCSIS) delineated by CableLabs. These core DOCSIS chipsets (CDCs) come from many different chipmakers, but because their functions are extremely standardized, not a lot of proprietary differentiation exists. In short, the CDC market is a commodity market.

Just because the CDC market is a commodity market does not mean it can't be lucrative, however. When an IC becomes non-proprietary, it's often because multiple chipmakers are eager to get their share of a surging market. The demand is usually so compelling that the torrential flow of shipping units overshadows the slimmer margins. And based on the technical merits of CDCs, that certainly could be the case, but it isn't.

CDCs can accomplish astounding things. On a standard, 6MHz-wide cable TV channel, up to 39Mbits/sec. of data could be downloaded using QAM 256. Even using QAM 64--the minimum allowable downstream encoding technique--27Mbits/sec. is possible. In addition, CMs can be shifted to different channels, so that the number of active users can be dynamically balanced among all of the DOCSIS channels. Alas, many multiple service operators (MSOs) only provide 1, 2, or perhaps 3 DOCSIS channels per fiber node. At best, that is less than 3 percent of the total RF spectrum available. However, while CMs only consume a sliver of an MSO's bandwidth, they account for large percentage of the revenue potentials. A typical MSO customer might spend $20 for basic cable, $15 for digital cable, $20 for premium channels and $40 for their CM service. Out of the $95 monthly bill, 42 percent of the revenue is coming from less than 3 percent of the bandwidth. With that kind of return, you'd think the MSOs would be anxious to spark the DOC SIS bonfire, but the opposite is true.

For example, CMs have DSL beat in terms of the maximum serviceable distance, maximum data rates possible and the capital expenditure required for service. But instead of pushing these advantages into victory, they throttle the data rates and charge just under DSL rates. Moreover, some are now trying to charge additional fees for users with multiple computers on a home network -- a move reminiscent of their failed attempt to apply a $2 monthly surcharge for every household TV. When they tried that, people got splitters. Now, they will buy NAT-enabled routers.

AOL Time Warner became a multi-billion dollar company because it learned early that the core of the Internet is based on community and sharing. It started with e-mail as its killer app, then moved on to instant messaging (IM). It promoted a concept of "build the subscriber base now, and the revenues will come later." If the MSOs would take a page out of the AOL playbook, they'd charge $30 a month for all the sub-net IP addresses you wanted and free, lightning-fast gaming servers within the regional area. Instead, like a tennis match in the middle of a war zone, they think their enemy is only on the other side of the net and don't see the massive threat coming over the horizon. Companies like SwitchPoint Networks and MeshNetworks are eager to replace DSL and CMs as the data connection of choice among residential users. And with their superior technology and the MSO's neglect for loyalty building, it is certainly possible.

In the end, the CDC market is a slightly shrinking revenue market, despite a 9 percent CAGR of shipment culminating with 16.6 million chipsets shipping in 2006. If the shipment CAGR grew to 15 percent, then CDCs would be a growing revenue market.

Eric Mantion is 6 senior analyst in the networking technology group at Cahners In-Stat/MDR.