To: Scrapps who wrote (7834 ) 2/28/2000 12:14:00 PM From: Terry D Respond to of 9236
DSL VS. CABLE MODEMS: OUTLINING THE FACTORS OF THE BATTLE SUMMARY: In this note we summarize and outline the key factors in what is perceived to be the "DSL vs. Cable Modem" battle. Our conclusion is that cable is the fat pipe with ultimately the higher bandwidth, which for regulatory and other bundling reasons should do particularly well in the U.S. In Europe and Asia, however, DSL's footprint advantage is more pronounced, and we should not underestimate the importance of the laptop market, which DSL should own. DSL also has better chip synergies with the 56K migration path. So far the cable companies have been more agile in quickly launching service, but this may not necessarily win cable modems the larger ultimate market share. HIGHLIGHTS: * In recent notes on PC-Tel (PCTI-$56-Buy), Aware (AWRE-$62-Strong Buy) and Westell (WSTL-$30-Buy), among others, we have been addressing some of our positions on the technology debates surrounding G.Lite and full-rate ADSL. In this note we more explicitly lay out our basic thinking on cable modems vs. various flavors of DSL. * First of all, we believe there is no doubt that both DSL and cable modems will be tremendously successful for the residential market. We believe that U.S. residential DSL subscribers will come from behind and exceed U.S. residential cable modem subscribers in 2001. * Ultimately, the global DSL opportunity looks to be much larger than the cable modem opportunity for the foreseeable future. There are around a billion copper connections in the world versus maybe a quarter as many coax connections. In addition, coax often do not pass businesses. * Coax will always have a capacity advantage over copper, in our view, especially if the world evolves from a broadcast model over coax to a switched model. It's like having a big cannon shoot-out, and in this shoot-out coax would likely always win if that is what the game is about. * However, in the big mechanized army vs. guerilla war, the big guns do not always win. Similarly, the sheer ultimate power of the coax pipe may not be the deciding factor in the bandwidth game. We argue that ubiquity will also matter, as will the evolutionary process from 56K connections. * In terms of ubiquity, not only is copper around the world with a billion homes vs. maybe 250m, but phone jacks are also in hotel rooms and in businesses. Even in any given home, there may be one or two cable connections, but 10 phone jacks. This means that coax connections will generally require more physical work and typically be less esthetically pleasing. * One important segment we believe must not be overlooked is the laptop segment. While it will certainly become possible to cram a cable modem into a laptop by 2002 or even 2001, we do not think this will be the broadband connection of choice for the mobile user. The killer device for this segment we believe will be a 56K/G.Lite modem, which automatically detects whether DSL service is available, but otherwise defaults to 56K. The mobile user cannot afford ever lacking connectivity, and the world will not convert to either DSL or cable modem service overnight. A 56K/G.Lite modem will also take minimal space, especially in "soft" format as per PC-Tel's invention. From the semiconductor perspective, the DSL-56K synergies are greater than any cable-56K synergies. * In the U.S. in particular, the regulatory and ownership situation will likely make cable into a very efficient competitor, especially as far as bundling and universal service requirements are concerned. This situation appears not to be as important outside the U.S. Warburg Dillon Read, LLC February 28, 2000 Anton Wahlman 212-821-3675