|
Gateway
access devices |
February/March
2002
By Tony
Morbin
|
Pace technology interview with Advanced-television.com
At the end of last year
advanced-televison.com's Tony Morbin interviewed Pace MicroTechnology's VP
of Strategic Business Development, Newtworks and Connected Devices Division,
Chris Boyce who provided a technical perspective on developments at Pace Micro
Technology.
ATV: "What's the future for the home gateway?"
Boyce: "The technological advances that are there to make, from a simple
Pay TV Box are:
First of all a high speed return path - for example our contracts with NTL and
Telewest in the UK and Comcast in the US.
"Secondly is the mass storage device - as used in the Sky box.
"Thirdly is some kind of home networking capability, of which there are a plethora
of standards and devices - and today we don't have any set top boxes where that's
built in as standard, but we have the gateway expander product which is designed
to connect to the set top.
"The fourth is a voice-over-IP processing engine.
"If you took the core product plus those four technologies you have a fairly
sophisticated gateway product. What's becoming increasingly clear is that the
adjustment in the financial community means we are unlikely to find an operator
in the near term to specify a product that contains all of those features as
it would be too highly priced.
"We had thought we would see the gateway expander technology going into the
set top box as standard by now whereas it looks like it will remain a separate
product to be made available to those who have expressed an interested in the
services it provides. It will be a supporting product.
"What we are seeing, is that you will end up with a more modular product range.
"Our customers are the network operators and the broadcasters and we are trying
to deliver our platforms that are then able to reach profitability as soon as
possible. There are two elements to profitability. First is enabling the right
services that you are able to charge for and get income from on a monthly basis
eg telephony, Voice over IP, high speed access for PCs, VOD, high speed return
path or a hard drive in the set top.
"Secondly you need to deliver those in as cost effective a way as possible,
keeping costs down.
"Our focus is on minimising costs associated with that platform, be it the price
of the backbone itself, reliability, auto-provisioning without use of call centres,
consumer installable equipment. If you have a modular system where someone installs
the core, ideally as you add functionality, you want to deliver that for installation
by the consumer rather than a service engineer going back. Pretty much all the
boxes we do from the simplest upwards have the capability for a new software
downloads."
ATV. "How do you see the PVR market developing? Is that what people will be
buying, because digital access will be taken for granted, with digital TVs after
analogue is switched off? Will the additional set top functionality be included
in the PVR?"
Boyce: "Its an interesting question about where the partition will lie
between what will go in a low cost set top, high end gateway and inside the
TV itself. I don't think that it is reasonable to ask consumers to replace every
TV they have in their house at analogue switch off. Therefore we need to have
ways to provide a box that will enable consumers to upgrade existing TVs to
the digital services. Between now and then there are issues of how the technology
migrates, from there being a very small number of digital TVs today to when
all TVs will be digital.
"The thing that worries me is the difference in product life cycle and the rate
of innovation in the TV market and the home gateway market. In three years at
Pace we've moved from downward satellite reception and dial-up return path to
digital cable boxes with high speed return path, there's hard drive boxes out
there. How much has TV changed in that time? We've gone wide-screen, but that's
it. I don't see the pace of change in the home gateway slowing down at all -
at least from the technology viewpoint.
"So, do you really want to burden the consumer with the cost of that expensive
piece of glass, with the need to buy a new expensive piece of glass every time
the technology needs updating? You also make things difficult for retail channels
if they have to stock four times as many TVs because one's got a hard drive
in and one's got a sky decoder in and another's got a piece of networking kit
in.
"The natural partition seems to me to be between gateway and television, much
as it is today, with the possible exception by the time you get to digital switch
on, if you have some more centralised gateway that perhaps has the hard drive
in it, and that is the core product and then you have peripheral screens about
the house.
"Then the way you get data from that centralised gateway to the end screen is
going to continue being compressed whatever the bandwidth and technology used,
whether its wireless or data over power lines or coaxial cable - whatever -
it is its going to be a channel of a limited bandwidth. Video being very bandwidth
hungry, its going to make sense to keep the video compressed to the point that
it reaches the screen. That means that at the receiving end, you have to have
some kind of decompression circuitry to display on the screen and you could
argue that that MPEG decoding capability might by then be cheap enough to be
built into every TV set.
"One thing to watch out for is that if the content that you are moving around
from the central depository, gateway and interface with the outside world and
say the TV sets around the house, is not free to air - its valuable content
- then not only will it have to be compressed, it will have to be encrypted.
Therefore at the receiving end you'll not only need decompression circuitry,
you'll need decryption, and particularly if you want to do some kind of interactive
things, which needs a return path as well, so pretty quickly you end up with
a set top again.
And again the question is, do you want to put that into every television? And
does that mean a Sky TV, a cable TV, etc - you are starting to multiply the
number of options.
"So I personally think that the right place to partition the technology, given
that its relative speed of innovation is the glass (TV) and whatever piece of
wire or wireless technology delivers content to the screen. By the time you
start to put cable modems in, and hard drives in, and maybe VOIP technology,
clearly you don't want to replicate that with every TV around the house - that
makes no sense.
"So the question is how can you make those services available to those other
screens, but only pay for one more expensive gateway? The answer is, yes you
can do that, but I would argue that you will still need a peripheral box rather
than something you integrate into the TV."
ATV: "And this is how you would handle the problem of legacy systems - existing
analogue TVs?"
Boyce: "Exactly. You have to manufacture a box to support legacy TVs,
so you end up with the scale of manufacturing those in their millions - until
its not worth putting (this technology) in the TV as the number that you sell
no longer justifies the cost."
ATV: "Is that essentially the product you are providing for digital terrestrial
access?"
Boyce: "The DTR500 can be a basic box without a modem, without conditional
access for free- to air programming, or it can be added in to be a high-spec
box - though its being marketed for the free-to-air programming. If a consumer
buys it they can then upgrade for Pay-TV."
ATV: "Whats the current state of play in terms of incorporating Voice over IP
in Pace set tops and how do you see the market developing?"
Boyce: "Its very exciting area. We are currently working on a product
called a Multimedia terminal adapter which combines cable modem plus voice over
IP processing engine, plus home networking.
"If you look at the cable industry internationally, everyone wants to offer
the triple play of video, voice and data. However the relative priorities of
those varies enormously. In the UK with just five free channels, TV is a very
important service that underpins that with high speed data and telephony following
on. In the US, if you look at the number of cable modems being shipped, you
could argue that data's one if the most lucrative elements.
"In places like Germany and other parts of Europe voice telephony comes first
with high speed data a very close second in terms of value. That's because there
are something like 40 channels of TV in Germany there already - and in terms
of German language programming there's only so much available (compared to the
UK where you can just take US product) unless you start dubbing programmes.
They have their soccer, and a wide mix of German language programming already
available to all. For a cable company, ability to deliver voice and high speed
data services over the network is the right entry point.
"Talking about this more modular strategy where you don't have to put in the
biggest box with everything in on day one, we are currently working on an MTA
product that has the ability to deliver voice and data services and can later
on use the return path for a set top."
ATV: "So is the MTA an add-on to the set top?"
Boyce: "No, it's the other way round. In the markets that will take it,
the MTA is the lead device which will be installed on day one for the voice
and data service. Then maybe there will people who want pay-TV as well and will
add on a set top. We believe that the MTA market is going to become an important
product line, because having a pay TV set top or gateway is not necessarily
the best way to penetrate a market where video is less of a priority than the
other two services."
ATV: "What sort of timing are we looking at for deployment, and where are
the target markets? Presumably Germany would be a key target market?"
Boyce: "Second half of 2002. Germany - Continental Europe, and to an
extent the US as well. What we've done is take our gateway expander and a VOIP
protocol using DECT, added a cable modem to it form our set top experience-
and that is essentially our MTA. Is a particularly good MTA because of the home
networking.
"If you are going to triple play services you've got to connect to the TV in
the lounge, a PC which is typically in a bedroom or study, and a telephone link,
typically in the hall, especially for multiple lines. If you are an operator
which is trying to support all three of those areas, physically you'd like to
put the device in three different places and you'd need to connect from one
to another, so the question is where do you put it and connect to the other
two services?
"Because we have the home networking capability built in using DECT, you can
put this device in the lounge - where a lot of European cable entry points are.
You can connect an MPA to it and then connected wirelessly via Dect handsets
and a PC with a minimal installation visit, and if they say they would like
pay-TV as well you are in the right place to deliver that. This makes it a strong
offer with low cost of ownership."
ATV: "What about people who want video as well from day one?"
Boyce: "If you are going to offer every single one of your subscribers
an option to take all three, the cheapest way to do that is to put it all inside
the same box. But then you burdening the cost of every box when a percentage
may not want a particular service. So then it comes to the business model of,
'How many people do I think will take pay-TV from day one?', 'How many later?'
and 'How many will never take TV?' We will design whatever the operator wants.
"The discussions that we are having today would suggest that in some markets
the modular approach is the right one. We take the sets of functionality and
bundle it to suit the particular markets."
ATV: "Although I appreciate that Pace will provide what the customer wants,
and when it comes to middleware standards, that is the middleware provider's
problem, the impression is that Pace does not favour use of MHP as a standard.
Is that a correct summary of Pace's point of view?"
Boyce: "Pace is on the steering board of DVB and very much welcomes open
standards.
"When you cost an MHP platform, and you compare one that is not MHP - in terms
of memory footprint, in terms of processor performance, you end up with something
that is significantly more expensive with MHP. Clearly when you are rolling
out millions of set tops and gateway devices, that's the variable cost element
of your business model, so any five cents you can save here and there multiplies
up pretty rapidly. Most people we have had discussions with have opted to go
the cheaper route.
"There are some technological trends that will affect this. There will come
a time when a 64 MB chip is the predominant chip used across the industry, and
that is sufficient to run MHP and whatever else and the memory burden disappears.
"I'm sure there will be MHP announcements and I'm sure we'll deploy some MHP
boxes, but at the moment you have to pay extra for MHP. I'm sure Liberate will
have an MHP element to their browser, and if the regulator decrees it they certainly
will.
"The PC is enormously successful, and the thing that makes it successful is
its compatibility. There's a big enough consumer base to make developing viable.
In the Pay-TV environment, if you are a Sky, there are enough subscribers to
ask a developer to build for that - platforms are being deployed and the iTV
is put on top, so it's a different type of rationale.
"The MHP argument makes life easier for the iTV developers. If you want to write
for satellite or cable you need to do a different ap for each but the deployments
are big enough to make each viable.
"In the pay TV environment it's a different model. You are presented - as a
consumer - with content and it's an 'industry' decision rather than consumer
decision.
"Our latest product is the DI6000, a product in development which we will deploy
with TV Cabo in Portugal. We are one of two set top box providers for TV Cabo.
Doccis Cable Modem, Microsoft TV advanced middleware."
Interview
conducted December 2001
