DSL Prime: SBC Without Spin

While some companies are planning for the capital intensive future, SBC is preparing a capital defensive future (but check out those nifty set top boxes), and MCI is becoming an attractive buyout target, with no plans for the future.

by Dave Burstein DSL Prime [July 7, 2004]

Mike Capellas's Side Deal with the Bells
Deal with New Edge Resells Bell DSL
"We're forecasting 20 to 25 percent down on consumer," Capellas told CIBC. I asked him whether he would do what Dave Dorman speaks of, build out COs across the country to extend the old Rhythms network with 700 active COs. "We want on-net calls," he reminded me, a reason he'd keep that alive, but clearly wants to avoid a facilities build. Capellas' further comment, that the company had "excess cash" that it was considering how to return to stockholders, simply doesn't make sense if he intends to remains much of a contender. Sounds like one less player in consumer wireline, as MCI winds down to a likely sellout as soon as the right bid comes in.

MCI still has millions of consumer relationships, and looks to cash them out by in effect accepting commissions from the Bells and Covad. New Edge has put together a remarkable network of all the Bells and most of the independents, reaching over 80 percent of the U.S. with DSL through 8200 COs. Whether they're considered a "master re-seller" or "systems integrator," their virtual network and extensive switching have enormous reach. This one-stop shopping made things easy for Capellas to choose. MCI has the technical strength to integrate a similar network, but is choosing to work through Dan Moffat's group instead. That avoids both investment and political complications.

SBC Without Spin
DSLAMs at cross-connects and other field units
Ed Whitacre announced an "SBC strategy could result in an incremental investment of $4 billion to $6 billion over five years." That includes "deploying fiber deeper into neighborhoods," and "testing an IP-based switched television service based on the Microsoft TV IPTV platform." "Fiber to the Premises (FTTP) for new network builds, such as developing subdivisions," would represent a miniscule fraction of subscribers, which I would guess would be fewer than 10 percent after a decade. SBC's position is "the cost, deployment time and customer inconvenience required for FTTP deployment in existing neighborhoods makes widespread deployment impractical for SBC companies." The release speaks both of " nodes within 5,000 feet of homes and businesses" as well as "nodes that serve 300 to 500 homes," which generally would be much shorter distances. The current network planning at SBC is oriented to the former; if they were headed to the shorter distances of 300 to 500 homes, many locations would be VDSL, not ADSL.

Simplifying, SBC is putting inexpensive DSLAMs in the field, with fiber backhaul. They will use ADSL2+, the current generation, meaning many customers will get 15 to 20 Mbps downstream, and one or two megabits upstream. This is less ambitious than Pronto, circa 1999, which deployed full DLC remote terminals, for voice, data, and even fiber. Pronto involved substantial new fiber builds, going to new locations. The new units will sometimes be closer to homes (5,000 foot goal), but will be far less expensive to deploy because they will predominantly use existing cross connect boxes and rights of way. This is possible because cheap 48 port DSLAMs, small enough to fit almost anywhere, can now be line powered. Backhaul may be Gig-E rather than ATM, with companies like ECI enthusiastically introducing small units with Gig-E.

Result: Perhaps 40 percent of SBC customers are close enough to COs and DLCs to get 15+ Mbps service with ADSL2+ today. Using more field units, SBC will be able to offer that speed to probably 70-80 percent. Whether the Microsoft TV deal is significant is unclear. The press inference from the announcement is a major shift in strategy from satellite video, but that's unlikely. However, IPTV for VOD and service beyond satellite makes a lot of sense. The industry still assumes SBC's primary strategy is satellite video with a DSL return, potentially a profitable architecture.

For SBC: Active elements, advanced set tops
Ross Ireland, as SBC's CTO who managed Pronto, told me "I'm never going to put active electronics in the field again." PONs reduce operating costs, one reason Verizon is going all the way home with fiber. Chris Rice, who recently replaced Ireland, instead is planning to put tens of thousands of DSLAMs in the field, dramatically less expensive in the short run. Rice is "maintaining the highest levels of discipline and responsibility with our capital investments." Folks at Verizon are also re-examining the PON question, looking at alternate fiber architecture that delivers more bandwidth but requires more switching.

SBC plans to make advanced home gateways a competitive advantage over cable. It's not coincidence that the DSL Forum has just published LAN management specs that are carefully designed to work well with Microsoft TV. Adding DSL to satellite gives a return path, essentially for home gaming as well as interactive TV. If the "gaming" becomes "gambling," money can be made. DSL at 15 Mbps is a natural complement to set tops with 300 and 600 gigabytes, which could store a hundred movies for instant play. Neither TIVOs nor set tops ship today with that much storage, but the designs I saw at Thomson and Broadcom include ATA and USB2 connectors allowing customers to add extra hard drives. Drive makers are predicting 500 GB drives under $100 in a few years.

All this requires new software, both for the gateway and management system. The DSL Forum TR's provide a framework that goes beyond CableLabs' CableHome. 2Wire has just lured back Hunt Norment from Broadjump to oversee their Component Management System (CMS), Siemens/Efficient has announced in addition to Tango, and Motive/Broadjump offered to brief me on their contributions to the standard.

Groveling to Microsoft is exceedingly uncomfortable for the Bells. SBC actively supported the Microsoft antitrust trial. But "we needed something that worked, and Microsoft seems to have it." The decision was made at the highest levels, with considerable internal dissent at one Bell.

Copyright 2004 Dave Burstein.
The DSL Prime Newsletter is reprinted with permission.

"The power of the printing press belongs solely to those who own the presses"
—A.J. Leibling

The Internet is the cheapest printing press ever invented.