[Tac Berry interview, a must read...]
*********************************************************************
Berry's Amati Pushes ADSL Market Forward
Source: Inter@ctive Week
Inter@ctive Week via Individual Inc. : A year ago, there was hardly any Digital Subscriber
Line, or xDSL, technology at the Comnet show. Now it seems to be everywhere. Unlike some
other new technologies, xDSL appears to be coming into the public eye as commercial
products are becoming available. Would you agree?
I think it actually came to the forefront way too fast. At the beginning of 1996, if you look at
what the stock prices of xDSL companies were doing, we got way out of control too quickly.
What happened was that an application became obvious to use this technology in Internet
access. What has never showed up for some applications, like ISDN [Integrated Services
Digital Network], was the application.
Amati was one of the companies that pioneered Asymmetric Digital Subscriber Line, or
ADSL, but now every data networking and telecommunications vendor in the country seems
to have an xDSL strategy. How do you confront so many new competitors, especially when
most of them are much larger than Amati?
I don't think we are going to confront those people. We'll partner with those people. We want
to be the ADSL enabler to folks in the remote access market. Some companies are saying,
'We'll wait for the semiconductor companies to develop the chip sets and we'll build it
ourselves.' That's fine for them.
But even big companies can't be everything to everyone, especially when you start talking
time-to-market. We can provide the ADSL solution, the board, the module and maybe even
the software solution, as we did with Texas Instruments [Inc., a licensee of Amati technology
to develop chip sets]. We are looking to partner and to speed up the market.
I think when it gets to the stage where companies are building network gear, like ATM
[Asynchronous Transfer Mode], then it's foolish for us to try to compete with major players
in that realm.
We want to focus on where our expertise lies, and we think we can do that and help move the
market forward. For other things, we'll be looking for partners as we did with Sourcecom
[announced at Comnet as a joint development partner with Amati for ADSL-to-ATM gear].
But you are also building your own equipment, such as the Allegro access multiplexer.
That's the next obvious question -- why build the equipment?
The answer is simple: We are building this equipment to help get the market started today. If
we don't do something to help the market get started today, then in 12 months we'll have to
have the same conversation.
The carriers needed some kind of networking equipment to enable them to get started, to get
the business going. If we need to help service providers do that, then that's what we'll do.
That's why we created the Allegro. We built it to match what our customers said they need
now.
Discrete Multitone, or DMT, is the North American standard for ADSL. But will DMT
modems from different vendors interoperate?
We've definitely tried to promote interoperability.
We have multiple chip vendors -- TI and Motorola [Semiconductor Wireline Group] for
ADSL chips and NEC Corp. for VDSL [Very High-Speed DSL], and now there's Alcatel
[Telecom]. Three will use our technology; one will license it.
Also, we will build a modem based on Motorola's and on TI's chips, and we will make sure
they talk to one another.
Of course, you have other DMT vendors -- Analog Devices [Inc.] Orckit [Communications
Ltd.] -- and we are working with them in the Interoperability Workshop. When they are ready
to test products, we will test products.
It really takes a direct testing of the products to know where the interoperability questions
may lie, but we can't do that until there are products to test.
There have been reports of some slippage in the delivery of the single-chip solution from
Motorola, due in March. When do you expect to have a modem based on that chip set?
We will have a Motorola-based solution in the second quarter of this year. There may have
been a little slip, but I don't think it was much.
And TI will have design samples this quarter that we can use to design firmware, and they
will have a product chip by the third quarter. We will have a product based on the TI chip
shortly after that.
Totally separate of them, we have been working on our own. For the Allegro, we have
miniaturized to the point that we can put two modems on a single card. When the Motorola
and TI chips are available we'll go to a quad card design with four modems.
Can the DMT solutions coming out this year compete on price with Carrierless Amplitude
Phase Modulation, or CAP-based solutions?
We were actually pricing our products as competitively as the CAP products that are
available now. We actually see a lot of price competition within DMT.
Alcatel won the contract from the JPC [Joint Procurement Consortium, consisting of
Ameritech Corp., BellSouth Corp., Pacific Bell and SBC Communications Inc.] with some very
aggressive pricing.
We understand Analog Devices is about to deliver a DMT solution that will compete on
price.
The truth is I don't see a lot of discussion among the carriers over CAP vs. DMT. British
Telecom [PLC] is probably picking vendors that are DMT-based, from what we hear.
It's the vendors that are talking CAP vs. DMT.
To date, the commercial deployment of xDSL has leaned heavily on High-Speed DSL, or
HDSL, and ISDN-based DSL, or IDSL. Would you say use of those technologies speeds up
the advent of ADSL in the market or slows it down?
I think it speeds it up. Using ISDL or HDSL -- that explains where this market is going.
Vendors provide the equipment that provides service today. [The carriers] need to be moving
ahead and starting to offer services.
If you start offering 128-kilobit-per-second service, like an IDSL offering could provide, you
can attract a solid customer base.
Then you come along with a 6-megabit-per-second service; that's enough of an incentive for
the customer to switch.
In the meantime, the carrier has built a sales force and a marketing plan, and they've gotten
the business off the ground. Moving to the higher-speed technology is the next logical step
for that business.
If they don't figure out how to do that, how to build the business, then to some extent it
doesn't matter whether they start with IDSL or ADSL.
<<Inter@ctive Week -- 02-17-97>> [02-20-97 at 17:24 EST, Copyright 1997, Ziff-Davis Wire]
|
