Upgrades Winding Down, Cable CTOs Look Ahead


Having completed most of the two-way upgrades to their cable networks, the chief
technology officers of the nation's major MSOs now are turning to their next set of
priorities: preparing for and launching advanced services; migrating to standards-based
cable-modem and set-top architectures; and planning for the reality of the retail-oriented
environment mandated for cable by the government. To gain some insight into new service
rollouts and the issues surrounding them,
Multichannel News senior broadband editor
Bill Menezes and Roger Brown, editor in chief of sister publication
Engineering & Design, interviewed Tony Werner of AT&T Broadband & Internet
Services, James Chiddix of Time Warner Cable, Alex Best of Cox Communications Inc. and Dan
Liberatore of Adelphia Communications Corp. An edited transcript of the interview follows:

MCN: What are your top three projects?

Chiddix: I think it's the obvious stuff: It's modem
rollouts and launching advanced services on digital set-top boxes and the OpenCable

Werner: The network upgrade as a general category
would be one. The [General Instrument Corp.] 'DCT-5000' [set-top] and the whole OpenCable
migration would be two. And the telephony launch would be three. Not necessarily in any
order of importance.

Best: Well I would say, putting on my cable-operator
hat, it is similar to Jim's: data-ready homes passed, telephone-ready homes passed,
digital-video-ready homes passed. Because obviously, the majority of our growth these days
comes from new services, we are trying to expand those footprints as quickly as we can.
Related to those three, we have lots of subprojects. [For example,] getting the backup
generators for the telephone service in place is a major undertaking for us.

Liberatore: No difference here. We've been very active
trying to get our rebuilds done and, as Alex calls them … getting the plant ready for
these services. High-speed-data services through the DOCSIS [Data Over Cable Service
Interface Specification] platform, getting those platforms out into the headends. And the
digital-TV platforms, trying to get some of the advanced applications figured out and

MCN: What's the status of your upgrades?

Chiddix: We're 85 percent done with our upgrades. We
will be done at the end of the year 2000, with the exception of recent acquisitions that
may still be in process. The vast majority are 750-megahertz upgrades, two-way. We've got
some 860 MHz, primarily in New York City. And we probably have some small stuff that's at
lower bandwidth.

MCN: Why do you have 860 MHz in New York?

Chiddix: First of all, the cost delta is very low.
There is huge density, as you might imagine. And second, there are 27 off-airs, and
there's potentially more demand for bandwidth there -- especially broadcast bandwidth --
than most markets.

MCN: What about everyone else?

Werner: By the end of this year, exactly 51 percent
of our plant will be upgraded and capable of two-way communications. We'll have about 54
percent that will either be 550 MHz, 750 MHz or 860 MHz at the end of this year. As we
move into next year, it should approach the mid-80s for both of those numbers. We bought a
bunch of plant that was 550 MHz and will stay at 550. But just about everything we're
building is 860 or 750.

The 860 MHz equipment that we are purchasing today has
output characteristics that allow it to be spaced at the same distance as 750, so you
don't have any additional electronics associated. You do pay a subtle difference for the
electronics itself, which works out to between $2 and $5 per home passed.

Best: For what I call Cox pro forma -- prior to the
recent four acquisitions of TCA [Cable TV Inc.], Media General [Inc.], Gannett [Co. Inc.]
and AT&T [Broadband systems] -- we're budgeted to be at 74 percent 750-MHz at the end
of this year and 67 percent two-way at the end of next year. The acquisitions probably
reduce those numbers. TCA was not as far along, obviously, on 750 MHz as we were. So I'm
still trying to get my hands around what the new numbers look like.

We have some 860 also, but actually, it's kind of strange
-- 860 can show up in some unusual places, like what we call our middle markets, like
Roanoke, Va., about to start an upgrade. They are going to 860. Cleveland is going to 860.
Quite frankly, it has to do with timing.

I mean, we essentially told our systems, 'Look, if you have
an upgrade coming up, if you can go to 860 with a relatively small percentage increase
over what it would cost to go to 750, then sure. Go ahead and do it.' And it depends upon
the status of the system at the time they are going to do the upgrade. For instance,
Roanoke was a 550 system spaced at 750, so going to 860 was relatively easy.

Liberatore: I feel like we're a bit behind here. We are
about 50 percent upgraded. For year-end, that ought to be another 5 percent. Reverse
two-way is in the 40 percent to 45 percent range currently. We expect to be done year-end
2001. I think we're doing 860 in [Los Angeles] and some of those places simply for the
same reason -- there are a lot of must-carries and that sort of thing.

MCN: What fiber-node architecture are you currently
using, and how closely are you watching new developments, like AT&T Corp.'s
'LightWire' project?

Chiddix: Well, we are continuing on the average to
have nodes of about 500 homes passed -- that varies with density. It's really based on
about a 1-kilometer radius around the node. That has to do with economics. In the headend
at this point, we are generally combining four of those together, and as traffic goes up
with more routed services of various kinds, it will be very easy to subdivide them back to
500 homes. And then it's also very easy -- we've got spare fibers -- to further divide
those 500-home nodes as needed.

Werner: We limit nodes to a maximum of 600 homes
passed per laser. And we are experimenting with a few pilots of the LightWire
architecture, which takes fiber down into nodes of about 70 homes passed, on average.

We won't be doing it everywhere, just because there's not
enough equipment out there to support it. But we think it's a real solid architecture, and
its advantages to us are reduced operating costs and fewer actives.

In Salt Lake City, we have a 520-mile area that is about
100 percent designed. Construction is about 50 percent complete, [with] several-thousand
customers connected to it. We are running 75 percent to 80 percent fewer active devices
per mile in the new architecture. The thing we can't quantify yet is what that means in
operational savings.

Best: We still are on the page of 1,000 homes max.
On average, our node sizes are 750 homes. We still take fiber down to those nodes in a
route-diverse path. We light up the diverse or the backup path when we launch telephony in
that market. We currently have 1 million telephone-ready homes where the backup path has
been illuminated, and we haven't had a need yet to subdivide a node.

We certainly have an interest in what AT&T is doing in
Salt Lake City, but even if they succeeded beyond their wildest dreams and went down to
100-home nodes, I am not about to go back and redo it. So certainly, we want to see what
the parameters and drivers are for that, but I am where I am. And I see no need to change
that strategy at the present time.

Liberatore: We believe that reliability is the key here, so
we build a node plus one amplifier and, occasionally, as we need, potentially, a third
line extender. But we do vary from that when we get into some of the underground areas. We
have seen some pretty small ones, from a couple of dozen homes to 300 homes. We did that
because we didn't want to go back [and pull fiber deeper] if we needed to do that.

When we did the analysis four or five years ago, the cost
delta was around 10 percent. It's gotten better because the optical prices have come way
down. As we get out into some of the other markets, I think we may have to back off of
that a little bit.

MCN: Any thoughts about AT&T's project?

Liberatore: Our amplifier counts are [about] two-and-a-half
or so per mile, as opposed to what I think is still maybe a four-amp-per-mile kind of
number. But our amplifier density is much lower, so I think the powering costs should be
lower and the operational costs should obviously be lower.

We think the reliability is there, but some of these
numbers are borne out in a level of uncertainty -- it's just the statistical nature of
trying to figure some of these things out. So I think we are saving money operationally.
But I'm not so sure, again, how far you have to drive this stuff.

Chiddix: I think it's going to be interesting to see
how this plays out. Historically, operating savings as a justification have proven
elusive. I remember there were fiber-to-the-curb models that were supposed to pay for
themselves based on operating savings. So I think it just pays to be careful about that

Alex is very much to the point. It is a matter of timing.
Clearly, optics get cheaper and so forth with time. If you start later, you can take
advantage of some later electronics. And maybe some better economics can maybe drive fiber
a little deeper.

Eventually, you run into the problem with getting into a
lot of underground. If you have to dig up a lot of underground to get fiber to smaller
nodes, then the costs begin to skyrocket.

But we are making careful measurements in our Rochester
[N.Y.] system, which is the kind of architecture I just described. And we are achieving
"four-9s" [99.99 percent reliability] there regularly in terms of availability.
And we are running a successful modem business, a successful telephony business and
getting digital boxes to work quite well. So I think we are quite satisfied with the

MCN: What's your migration strategy as it relates to
DOCSIS and OpenCable set-tops?

Chiddix: All of our new launches are DOCSIS, and we
are beginning to look at rolling out DOCSIS in parallel with our legacy products. I think
we will have some legacy products in place for years to come. Some of the legacy stuff
works quite well, although we clearly want to move to DOCSIS to take advantage of retail
and some of DOCSIS' more advanced features.

OpenCable at this point is all about retail. And we expect
to have POD [point-of-deployment] modules from our vendors available next July. We expect
to have set-top boxes with IEEE 1394 connectors on them next October. And we are very much
involved with the OpenCable middleware initiative, which will take a little more time.

So realistically, we are a couple of years away from
retail, OpenCable devices becoming a big factor in our business, at the very least. But we
are laying the foundation for that. In the meantime, like everybody else, we are launching
proprietary boxes and keeping doors open so that they can coexist with the OpenCable
devices to come.

Best: Any new modem launched has to be … a
DOCSIS product. We have selected two vendors' CMTS [cable-modem-termination systems]:
Cisco [Systems Inc.] and Motorola [Inc.]. On the modems, as far as what we've purchased,
DOCSIS-based, we've selected GI, Toshiba [America Consumer Products] and Samsung
[Telecommunications America Inc.].

We've also requested that all of our systems that had
launched proprietary modems transition to DOCSIS, and in fact, I think it's complete. We
will have no need to be buying proprietary modems before long, as we will have
transitioned all of our systems to DOCSIS.

As far as retail sales, it's a mixed bag where our systems
are with respect to the retailers in their markets. I would characterize it in general
today as early phases of working with retailers on a retail strategy.

As far as OpenCable is concerned, to be blunt, we've done
what we legally have to do, and that is place orders for PODs. Quite frankly, we'd be
quite surprised if we see any host devices to put those PODs into in the retail stores
this year. I think it's a little premature to determine exactly what our strategy is going
to be with respect to OpenCable.

Obviously, we have to support retail devices. There is some
question in a lot of people's minds as to whether the consumer will really go to Circuit
City [Stores Inc.] and pay … You know, I'm buying them, let's say, for $300. I assume
that Circuit City wants to make $100. So it's not obvious to me why someone would want to
buy a $400 set-top, and then have me plug in a $75 POD. Yet, versus leasing the set-top
from me -- especially under the auspices of what I call national equipment averaging,
which we are using, which means that when you get a digital set-top from us, gee, the
monthly lease fee might be $1.75 -- I don't quite understand why someone would pay $400 to
replace a $1.75 lease cost.

MCN: Do you think the retail side is more likely to
develop when integrated digital TVs start showing up in a few years?

Best: Yes, I guess I could go along with that. I
think from a cable-industry perspective, OpenCable and the concept of retail sales are
problematic from the standpoint that you buy a TV set and you probably keep that thing 10
to 15 years.

There's a great concern I have about how you build devices,
OpenCable devices, into TV sets that you intend to last for 15 years when I intend to
launch new services on an annual basis. Quite frankly, I don't know how that problem gets
solved, short of there always being a need for set-tops. I'd love to hear anybody else's
input on that, but I think that is the dilemma we all face in the long run.

Liberatore: All of our new launches are DOCSIS. We are not
buying any proprietary or legacy high-speed-data equipment. The legacy systems that we
have in place also have DOCSIS sitting beside them, so that conversion is getting made,
and I must say, from our perspective, we think DOCSIS is operating [with a lot more
stability] than what we have seen in the past.

Retail, I just don't think that we can take a position on
that. Although we are really trying to make sure that we don't get ourselves in a position
where that becomes a problem. Now I'm not so sure what that is. I agree with exactly where
Alex is. The thing that concerns me most is how we try to manage the quality of the
product and the applications now that they can come from multiple sources.

So we try to understand, and we try to keep that on the
horizon. We make that part of all of our discussions with the current vendors we are
dealing with on set-tops, but how that's going to come out, I just don't really want to
try to piece that together at this point.

Werner: We emphatically believe in the retail
channel for set-tops for a whole bunch of reasons. When consumers go in to buy a new TV
set today, they always get pitched on the products that are there. If cable's not there,
we lose that opportunity. It is absolutely critical that we have a product in there that
is able to be sold. So you will see us getting more and more aggressive on that front.

I think even GI and Scientific-Atlanta [Inc.] want a retail
route. And lots of other people like Thomson [Consumer Electronics], Philips [Consumer
Electronics Co.], Sony [Corp.] and Matsushita [Consumer Electronics] also want to sell
product. But we have to get in there, start getting integrated with other devices and have
a product on the shelf.

MCN: How do you guys feel about trying to integrate
multiple pieces of software together in digital set-tops and the potential problems that

Chiddix: Certainly it's technically challenging.
It's not technically impossible. The issue is what we are trying to accomplish. If we
truly want to have appliances -- and I agree with Alex that set-top boxes are not going to
sell at retail -- but if we are going to have TV sets that work well on cable as they come
home from the store, we have a challenge that our competitors don't have.

A DBS [direct-broadcast satellite] company has a nationwide
footprint. Therefore, a DirecTV [Inc.]-ready TV set will work anywhere in the country. If
cable is to have that kind of portability, then we are going to have to have some kind of
common software interface. So no one is claiming it is easy, but the computer folks who
have dug into it really believe that it can be done if the industry really wants to make
it happen.

Best: [Laughs] Let's see if I can be diplomatic
about that. You realize that you're asking a guy who has suffered through a lot of issues
related to integrating just two pieces of software on a box: an operating system and a
guide. And struggling with making sure that the two versions of those pieces of software
are compatible and don't get out of synch with each other.

So obviously, it concerns me if that number of software
platforms grows to three or four or five, with each one provided by a separate company,
each one constantly adding revisions and upgrades, and I am sitting in a headend trying to
make sure that five pieces of software with multiple revisions are compatible with each
other once they get loaded in a box.

Like Jim said, not impossible, but a concern and
problematic. And it just means that we are going to have to put testing facilities in
place to absolutely ensure that once we provide upgrades and revisions to any of those
pieces of software, they remain compatible with the other pieces.

Chiddix: We've had an interesting warm-up exercise,
though, in DOCSIS, and I certainly would agree that DOCSIS is much, much simpler than what
we are talking about. But I think that DOCSIS has been a real accomplishment in terms of
being able to give customers assurance that if they buy something at retail with the
CableLabs-certified [Cable Television Laboratories Inc.] sticker, it's actually going to

We are going to need to do that in spades in terms of truly
cable-ready digital-TV sets that are application-ready. In other words, TV sets that are
ready for VOD [video-on-demand], and so forth.

Liberatore: I guess I agree with both Alex and Jim. Nothing
ever goes as smoothly as our vendors or our government would like us to believe it will. I
do believe that CableLabs' efforts, and Jim used the DOCSIS example, are a fine example of
showing how this can get done. I think it's probably an order of magnitude harder than
DOCSIS, but we probably have an order of magnitude of resources at this, and I think we
will get it done. It's just certainly not going to be a smooth journey, and there are a
lot of traps.

We are not going to be able to rely on our vendors as much
as we used to, to bring these things together for us. As Alex said, we are going to have
to do some of this integration ourselves. We are actually investing in some of the tools,
which we have not done as a company before, to start to bring some of these things
together for us and bring some of this effort internal.

MCN: Does that mean you're hiring software engineers
and developing a test facility?

Liberatore: Yes. We are working on some things internally,
at least in the outline that we've talked about. We've put an IT [information technology]
group in place that is for advanced development of these kinds of things. So I think we
are getting positioned to do a better job with that.

MCN: What are some of the advanced services you hope
to roll out, and what is their status?

Liberatore: One of the things that our advanced product
group is working on is the back-office stuff, so that you have a lot more flexibility in
how these services get presented and billed and how you bundle them. All of the obvious
interactive set-top discussions are going on. We also have a group working on e-commerce
and some of those things. It really takes a lot of effort to prioritize which [things] you
want to deal with.

Best: Once again, Cox pro forma, before the
acquisitions, we only had 18 systems. We've launched digital video in 15, Internet in 14
and telephony in seven. So we are going as fast as we can hooking up new RGUs
[revenue-generating units] at a clip, I believe, of around 40,000 per month of those three

And we have two new trials budgeted for next year, both in
the second quarter: an Internet-to-the-TV trial through a digital set-top over the S-A
platform, and a VOD platform, once again on the S-A platform. But certainly, with respect
to VOD, it's clear, I think, that Jim is the leader there in terms of those trials.

Liberatore: Let me just add that we have a small VOD trial
going on, and we have a small e-mail to the set-top trial going on, and we have a small
HFC [hybrid fiber-coaxial] telephony trial going on. Nothing to the point where we could
start to boilerplate rollouts with those things, but I think there is a little bit of all
that activity going on with us. I wish I was as far along as Alex was with telephony.

Werner: We are moving ahead with the telephony
initiatives, as planned. We have eight markets so far that we have equipment installed in
and are testing, and at least three or four where we've stepped into commercial launch of
the product.

We have very, very high levels of performance. Right now,
we are testing every line as we install it, and we test every one of the incumbent's lines
right before we disconnect it. On just about every possible measurement, we offer a
higher-quality product. So we are enthused based upon that, and it's now just a matter of
getting installations done, getting to scale and starting to move ahead.

Chiddix: We have three VOD sites up and running.
They are not ready for paying customers yet, but they work. We are rolling out digital
set-top boxes at a great rate. We should be well above 400,000 or 500,000 by the end of
this year, and we may be heading for a couple of million next year.

That is a big deal operationally. When that is under our
belt, we hope we will have VOD ready, which is, we think, the big revenue opportunity
near-term in interactivity. And I think we have 250,000 or 270,000 [data] customers as of
this week.

In terms of telephony, we have our telephony project up and
going in Rochester, and it works fine. Still, it is not as interesting a place to put
capital in as modems and digital set-top boxes are at this point.

MCN: Could you touch on the impact that
consolidation is having on operations?

Best: Well, in the existing systems, I would say to
a large extent that it's somewhat invisible to them, other than if some of these
acquisitions are in fact in close proximity to those systems, where the two have to be
merged together.

That is, in fact, the case in Oklahoma City -- we owned the
doughnut hole, and Gannett [Multimedia Cablevision Inc.] owned the doughnut. And we bought
[Multimedia]. So Oklahoma City now has the task of integrating two cable systems together,
and obviously, that takes a tremendous amount of effort to consolidate the teams, to get
the platforms similar.

And there's always the culture issue that has to be dealt
with. As far as the stand-alone acquisitions, it's trying to decide whether we want to
keep the same management team, or whether the easiest way to change the culture is to
indoctrinate some of our people into those acquisitions. We have to go in and take a look
at their platform and decide if it's capable of providing the kind of things we want to
do, and if not, we have to figure out when to start a design, when to start an upgrade,
how do we manage our capital expenditures related to that activity and how do we train
those people for our ways. So it's not easy, and it takes a lot of time away.

MCN: One of the big issues surrounding rolling out
new services quickly is auto-provisioning. What are each of you doing in this area?

Chiddix: We are beginning to do some testing of
self-provisioning. There was a big debate a few years ago about whether people should use
high-pass filters to clean up their cable systems' ingress, and at the time, we felt that
was shortsighted, and I think that is being proven out.

We are winding up with the majority of our homes taking
two-way services. Therefore, the filters just postpone the inevitable cleanup process, and
they get in the way of self-provisioning. It's hard to self-provision when you have to
take a filter out when someone takes their cable modem home.

Best: We will be testing that in other systems. We
filtered most of our systems, with the exception of Las Vegas, which we purchased, and
Omaha [Neb.], which chose not to do that. But we don't see that as a big problem. We
remove the filter when we install the additional outlet for the service, but we think also
that self-provisioning will be a tremendous boon in terms of the rate at which we install
the new services.

Liberatore: We've used filters as kind of troubleshooting
or Band-Aid approaches until we can get into the house to solve the problem, but not as a
rule do we filter.

We have an auto-provisioning package that we are going to
start to test. Obviously, for all of us, the installation process for high-speed data has
probably been the most cumbersome thing, so we are trying to get that auto-provisioning
package out to the customers and test that.

Werner: Everyone is working on auto-provisioning,
but it's just one piece of the puzzle. Getting to the point where the initial install is
done by the consumer is probably a little ways off, though. That is part of the
home-networking effort that CableLabs is doing, that we're doing to try and get there, but
it's one that's not real easily solved.

The side-of-the-house [approach] with the 'Packet Port'
idea that's got four lines of phone and one line of data is great for that. If [a consumer
want to change service], all you have to do is call us up, plug it in and away you go. We
don't roll a truck for that. But having a consumer do his own install on a primary line is
probably not on our near horizon.

MCN: What's the big obstacle in achieving the
success you want to enjoy? Second, what do you think the world will look like 10 years
from now?

Chiddix: I think our big obstacle is just
impatience. We've got growing competitive [pressures], of course, but I think we've laid
the groundwork with the upgrades and with the digital technologies that are available to
be a very competitive player and to get into a whole bunch of new businesses. I'm really
optimistic about all of that.

Looking out 10 years, I think we will still have an analog
tier. I think one of our strengths is full-house service to analog TVs, and there's a
whole bunch of them out there, although I think our analog tier may be much smaller than
it is today.

I think in 10 years, VOD will be a huge business, including
high-definition VOD. I think the consumer folks will overcome the barriers to making
great-big cheap displays, which is the missing piece for HDTV to take off. And I think
we'll have a very robust tier of MPEG services, from VOD through broadcast, standard- and
high-definition, and we'll have a very robust tier of IP [Internet protocol] services.

Best: The hurdles? People. I think we have openings
for 600 people right now. Obviously, with low unemployment in many of our markets and the
competition that exists for telecommunications services, it's very hard to find qualified
people to do the businesses that we want to do.

[Looking ahead] 10 years from now, what I'm excited about
is VOD. Although there are many intellectual-property issues to be dealt with, I think
that technology will enable you to sit in your home and potentially have access to
libraries around the world.

Strangely enough, 15 years ago, there was a company at a
cable show called World Video Library. I think they were about 15 years before their time,
but I think that is what the customer will have access to -- the world video library.

Liberatore: Clearly, it's execution and personnel for us.
Our products are more complex and they are more varied. So getting the people in place to
get these things out there and execute like clockwork over the next couple of years is
really the key that we struggle with everyday.

[Looking ahead] 10 years from now, there's going to be a
wide variety of products. I don't know that I want to define those products, but it will
be a wide variety of products from us, and I think they'll be bundled as the customer
wants them.

And the Internet has really changed the way we get
fulfilled on our products these days. I believe the customer should be able to buy what
they want, when they want it and wherever they want it, and they ought to be able to get
it from us.

Werner: I think two or three things are becoming
real clear to me today. Everybody will have broadband to the home. The home will be
always-on. There will be a variety of IP addresses in your house, a variety of smart
devices, all of them taking in information off of this broadband network, and all of them
giving information back to it, as well.

Just about everything that can be turned into a byte will
be turned into a byte. Look at MP3 [the downloadable digital-music format]. The quality
difference of it versus a totally uncompressed CD [compact disc] gets smaller every day.
If you load them onto a PC, it's the most convenient way there is to store, archive,
select and play music.

The same thing will happen [with] digital cameras. Even
professional photographers are starting to say that the multimegapixel digital cameras are
pretty darn good. And again, they are so convenient, and the quality gap is getting
smaller, and you can manipulate and store the pictures.

The same thing will happen with books and other media.
Hardback books will be collectors' items, and they will be a form of art and decoration.
If you want information, you'll download it to a portable device that will have a big

So I think everything will be digital, everything will be
broadband, everything will be constantly on and constantly communicating.