In Re: ) ) FCC BANDWIDTH FORUM ) Volume: 1 Pages: 1 through 208 Place: Washington, D.C. Date: January 23, 1997 Before the FEDERAL COMMUNICATIONS COMMISSION Washington, D.C. 20554 In Re: ) ) FCC BANDWIDTH FORUM ) Room 856 FCC Building 1919 M Street, N.W. Washington, D.C. Thursday, January 23, 1997 The parties met, pursuant to Notice, at approximately 9:00 a.m. BEFORE: REED HUNDT Chairman, FCC APPEARANCES: On behalf of the FCC: MR. REED HUNDT, Chairman, FCC COMMISSIONER NESS MR. ELLIOT MAXWELL, FCC MR. STAGG NEWMAN, Bellcore MR. ROBERT PEPPER, FCC MR. KEVIN WERBACH, FCC APPEARANCES (Continued): Additional Appearances: PANEL I: MR. JOHN CURRAN, BBN MR. DOUGLAS MORGAN, Winstar MR. DAVID REED, Cable Labs MR. LES VADASZ, Intel MR. PAT WHITE, Bell Atlantic PANEL II: MR. LEE BAUMAN, Pacific Telesis MR. KEITH FOX, Cisco MR. MATTHEW KORN, America Online MR. JAMES LOVE, Consumer Project on Technology MR. MIKE TREST, ATM Net PANEL III: MS. GLEE HARRAH CADY, NetCom MR. PETER HARTER, Netscape MR. DANIEL WEITZNER, Center for Democracy and Technology APPEARANCES (Continued): Additional Speakers: MR. DAN BERNENGER, Vocal Tech MR. ALAN CIAMPORCERO MR. ROY DAHNKE, SETA Corporation MR. KEN DEUTSCH, Issue Dynamics MR. RUSSELL DOYAN, Saylor MR. IHOR GOWDA, Newbridge Networks MR. GORDON JACOBSON, IRAP Network MR. DAVID LaPIER, Cisco MR. JOHN MAZER, General Datacom MR. BROCK MEEKS, MSNBC MR. STEVE NEVAS, Cyberbridge Associates MR. CHARLES ROSS, President, Canthos Corporation, BNET MR. BOB SCHMIDT, National Digital Network MR. ANDREW SHAPIRO, Twentieth Century Fund STEVE STEWART, IBM MR. JEFF WALDHUTER, Executive Director, Nynex Science and Technology MS. GWENDOLYN WOODARD, President, Worldwide Education Consultants I N D E X VOIR WITNESSES: DIRECT CROSS REDIRECT RECROSS DIRE None. E X H I B I T S IDENTIFIED RECEIVED REJECTED None. Hearing Began: 9:00 a.m. Hearing Ended: 3:40 p.m. Recess Began: 1:15 p.m. Recess Ended: 2:19 p.m. P R O C E E D I N G S MR. PEPPER: Good morning and welcome to the -- okay. Now I can say good morning. We are, by the way, putting this broad band forum online with both video as well as audio on the Net. There will be demonstrations of this beginning at noon down in room 845 that you can see. I'd just like to welcome everybody to the FCC's broad band forum and very quickly turn it over to our Chairman, Reed Hundt. COMMISSIONER HUNDT: Thank you all very, very much. I'm very pleased to see you and I'm especially impressed by the very distinguished panelists that we're going to have over the course of the day. You never know when you say that something is a first, you never know if that isn't a tremendous criticism of the fact that you didn't do whatever it was a lot earlier. Or if instead it's just a self-complement about the fact that you've had some kind of epiphany and you finally recognized that that thing that is a first is very important. I don't know whether this is an epiphanic moment or an appropriate moment for self-doubt and criticism for the FCC. Maybe it's a little bit of both. We have never had a forum like this. It's David Reed's fault really. He told us to have it a couple of years ago, but he wasn't persuasive enough when I first got here. I'm just kidding, David. It's great to see you. We've never had a forum just like this, but we have for the last couple of years been increasingly concerned at the FCC about what we call the problem of access and band width. We're not 100 percent sure exactly how to state the problem. We're not exactly sure what are the dimensions of the issues here. We have some grasp that public policy is becoming increasingly important to the emerging industry of band width delivery if I can call it that. But part of the purpose of this forum is to define what are the interactions between conventional telcom public policy and the band width issues that increasingly are preoccupying so many of the industries of the world, but particularly now as never before are an important part of the function of our traditional telcom industry. So I think what you're helping us do today is define the questions. Now, because we're dealing with communications, actually you can already look the answers up in George Gilder's article, for example. In many ways, we're just trying to take George's answers and those of you who like me have enjoyed reading George's answers for many years, you will understand that the problem with George's answers is just making sure that you do have the question matched up correctly with those answers. George has pointed out, for example, that we should not have the options of the spectrum because the spectrum as a band width possibility is now infinite and that spread spectrum technology means that we have no band width issue with respect to spectrum. $24 billion has been paid by people who don't seem to agree with George and who desire to have exclusive licenses. Possibly today you all will tell us that those people are all wrong. If you do say that, try to be just a little low key about it would you please? Because we have a number of other options coming up and Congress in fact believes that the balanced budget depends on George being wrong, at least in the next couple of years. I've also read Economist's answers to the questions that I hope you all will frame today. The Economist announced that time and distance have no meaning in telecommunications. This came as a very welcome relief to those of us working on interstate access reform and we suggested last night to a number of people in the industry that meant time and distance have no meaning and intrastate and interstate therefore is a concept that is of no relevance. This is a little bit of a problem to the world as we know it. $23 billion in charges to the telcom industry based on the definition of state lines, something that's actually part of our constitution. The Economist being published in Never-Neverland was kind of indifferent to the taxation system that in fact is built on the notion that time and distance have if not meaning a presumption of meaning. What are we supposed to do in short about the real world in which it may be true that in many technological ways time and distance have less significance than they ever had before in terms of communication, but it is not at all the case that they have no meaning in terms of the reform of the systems that currently exist so as in fact to provide a climate in which we can create the maximum incentives for the proliferation of band width and for the proliferation of access. I also read I guess it was last July that Bill Gates and Andy Grove had a fascinating dialogue in which Andy -- and Les is here to explain this today I hope. Andy said, if I remember correctly, the following. And I'm remembering this by reading the quote. "As exciting as the Internet is though, there's one big problem. Telecommunications bad width. And Gates responded, band width bottlenecks. No question. That's the biggest obstacle to where we'd like to take the PC." We actually found this particular colloquy to be immensely interesting and we are curious to know whether it is a synonym for a colloquy that we are having here in public in a variety of proceedings. In our colloquy, we talk about the exclusive facility bottleneck problem in communications. We talk about whether or not the last mile of the local loop, the spread of wire from the consumer to some centralized aggregation point. We talk about whether or not that is a bottleneck. We talk about that in antitrust terms. We talk about it in competition terms. We also talk about whether the current local exchange market is a bottleneck. That is the key issue in access reform. When we're having that conversation, what we'd like to find out today is are we also having a conversation on the same topic that Andy and Bill were talking about. Do they mean the same thing that we mean? If we don't mean the same thing, let's find out the difference. But if we do mean the same thing, then it should be the case that this obstacle to where we'd like to take the PC in fact is the topic that we are spending hundreds, indeed thousands of hours on right here at the FCC. And so now we've discovered that maybe our traditional public policy pursuits are directly related to the future of our information economy in an even bigger way than we may previously have thought was the case. The FCC for many years has had a group called the Network Reliability Council. It's primarily been run by the incumbent telephone companies. Although it's been open to participation from all the other industries that have ever wanted to be involved and many have wanted to be involved. The Network Reliability Council concerns itself with, among other things, failure rates for our communications system. The Network Reliability Council is the principle way in which the telcom industry creates the protocols that allow it to use redundancy in the network so as to make communications continue to work in the event of, and this has all happened while Rochelle and Susan and I have been here, earthquakes in California, hurricanes in South Carolina, numerous other calamities. We don't like to spend a lot of time dwelling on this, but the fact is there are dozens and dozens and dozens of incidents of this kind every year of various levels of magnitude. And when we say that we have the greatest communication system in the world, what we mean in this country, among other things is that our system does not fail in the event of these crises. Now, one thing that we're very curious about is why does my 11 year old get a busy signal when he tries to get on America-On-Line? And is that part of what ought to be the mission of the Network Reliability Council? Or in fact, should it be the case where we say, no, it isn't important that the Internet be reliable in the same way as the circuit switch network? Is that really what you all want to tell us, that isn't important? Remember that the Network Reliability Council is not guided by regulations. It is an intra-industry group that we bless and participate with. But fundamentally, it is a means for cooperation that allows everyone to take advantage of the redundancy that is built into the circuit's network. So one of the things we'd like to find out is whether or not reliability, certainty, the ability to utilize redundancy is in fact an appropriate policy issue with respect to the packet switch network? Or would you say to us, no, that's not of any particular importance and if I happen to rely on packet switch technology to get my sports or news or make stockmarket trades? Well, that's not terribly important if that doesn't work for several hours or days. Hmm. Don't misunderstand us. We do not in any way intend in this forum to be empowering government to do that which it does not need to do. Please understand the opposite to be the case because the opposite is true. What we would like is to have help from all the relevant industries, users, consumer groups, participants in our information economy. We'd like their help in figuring out the traditional telecommunications policy problems to the degree that they have any bearing on what I read in all of these articles is the fundamental issue of building the information highway by providing more bandwidth, cheaper access and greater availability. So I'm terribly, terribly excited about this forum because I think that it will in fact lead to that kind of help and it will permit us to get to what we're ultimately looking forward to and that is a profoundly deregulatory era in the communications system. Right now in this country, there are 10,000 people employed in regulating the telecommunications industry. There's about 200 here and 9,800 in the states. There are approximately 25,000 pages of orders and decisions that regulate the communications industry. There's about four dozen pages here and all the rest are issued at the state level. None of us, not our friends at the state level and none of us here at the FCC, desire to have this level of micromanagement be a necessity or even be desirable as the communications era unfolds. But none of us should fool ourselves. All of that requires a lot of change to do away with or to alter or to modify. So when I say we need help in effecting this change, I know you'll have to believe me and it's your help that we're counting on getting today. Thank you all very much. [applause] MR. PEPPER: I believe Commissioner Ness has a few words. COMMISSIONER NESS: Thank you, very much. That was a terrific introduction to the problems that we are looking at globally as well as domestically. And I'm delighted to have the opportunity to join everyone here today to explore some of these issues. We've reached a new stage in the long standing debate between the telephone companies and the Internet community. For too long, the argument has been one of whether enhanced service providers should pay access charges. The telcos say that the ESP exemption should be ended. The ESPs say they aren't carriers and shouldn't be required to pay carrier charges that aren't assessed on other users of the telephone network. A recent MPRN on access charges and the forum today represent a new stage in that debate. The discussion shouldn't only be about money, however, but it should really be focused on technology. The needs and expectations of computer users are different from those of consumers using the telephone network only for voice calls. We need to understand better how networks might more efficiently carry voice and data traffic, how new or different equipment and technologies could enable telephone companies and other service providers to better serve computer communications. And what costs are associated with existing and new ways of transmitting computer traffic. Our inquiries should go beyond the traditional wired voice network and consider non-traditional providers such as licensed and unlicensed wireless operators and equipment. We must also explore how increased competition in the provision of transmission services can alleviate the concerns we've heard, both from the telephone companies as well as from the computer industry. The more that we can rely on market forces to determine both the technologies that will be employed and the means by which costs will be recovered, the better will be served Congress's goals of competition and deregulation as the Chairman so beautifully stated. I want to congratulate the staff who put together a stellar program and as a computer user myself with a dedicated ISTN business line at home, two children fighting for time on our family's information service provider and my husband being online with his office interfering with my trying to send faxes, I'll be personally interested in seeing how the solutions are formulated today. Moreover, I look forward some day soon to following the proceedings in my office on my computer as many people will be doing today thanks to the FCC's online capabilities. So thank you very much and I hope it's a productive day for us all. Thank you. [applause] MR. PEPPER: I'd also like to recognize Commissioner Chong who's here. So you can see that there's a great deal of interest among our commissioners on these questions. I was reminded by our transcriber, I forgot to introduce myself. I'm Robert Pepper, Chief of the Office of Plans and Policy. I'll be joined today by Kevin Werbach who really, from our office, who really deserves kudos for putting this program together and Elliot Maxwell, our Deputy Chief of the Office, as well as other staff here from the Common Carrier Bureau, Wireless Bureau and other Bureaus. As I mentioned earlier, the forum is going out live on the Internet in both audio and video. Beginning at noon until 5:00 o'clock we have demonstrations in Room 845 of a variety of new band width technologies including ADSL. We have a demonstration of the direct PC, MMDS access and some other examples. And also the transcript of the proceeding today will be put into the record in the access charge proceeding that Commissioner Ness referred to. So this will become part of the formal record in that proceeding as well as the NOI, the Notice Of Inquiry, that specifically relates in the proceeding, that docket, to the ISP and Enhanced Service Provider questions. We'd like to make today informative and also interactive. We will start with a presentation of the various bandwidth technologies. We'll then move into panels that are going to be discussion oriented. We'll be asking questions. There will be opportunities for those of you in the audience to ask questions. And we hope that the panelists also ask one another questions since this is an area in which there are no fast and hard answer, but lots of questions which we all need to learn about. Our first presenter this morning is Stagg Newman. I'm missing page one of his bio. So I'll do it from memory. Stagg, as he told me earlier this morning, received a totally useless Ph.D. in mathematics. But since then, since the mid-1970s has been first I guess at Bell Labs. Stagg is responsible for work on fiber optics now with Bellcore. Fiber optics, wireless communications, battery technology, other band width technologies. Battery technology is really fascinating. Last night he came in and he showed me their new battery for his cell phone. He gave me the battery and it didn't weigh anything. And he showed me some plastic and it's twistable. It's very light. And from my perspective it's a radical breakthrough in battery technology. So Stagg is in the applied areas at Bellcore working on all kinds of new technologies that are going to help all of us in the communications industries. But this morning he's going to walk us through really a technology talk about these various bandwidth technologies and try to help us get a better understanding of what the test of the day is going to be all about. Stagg. MR. NEWMAN: Thanks, Bob and Elliot. I'd like to thank you for inviting me here and basically -- I don't have my glasses on. So I can't see which side says on and off. Thank you for inviting me. What I'd like to talk about today is the bandwidth bonanza and basically there's no question in my mind that there will be a tremendous band width bonanza for corporate America. We will see the same type price performance improvements we're already seeing, that we've seen in the computer world. That is a doubling every -- you can argue how many months, 18 months, 12 months, 24 months, of the price performance characteristic for obtaining bandwidth. But what happens to the mass market? The mass market is the key issue I think and that's a real challenge for all of us, both from a technology side, from a regulatory side and from an operations side. Said another way, Bill Gates was quoted recently over the Internet, of course, or misquoted, as saying in the year 2000, 90 percent of all the users of the Internet will still access it over dial up modems over the telephone network at 28.8 kilobytes per second or in reality usually less for those of you who have been on certain networks recently. Unfortunately, our analysis is he's probably right. Where are we today? Today it's estimated worldwide there are about 30 million users of the Internet. So we have this large wave overtaking the communications industry in ways that we can't understand. What I want to do is put a context setting in what I think the evolution of the Internet means. Basically replacing the universal service we've had today, dial tone, with a new universal service called IP dial tone. And I think for, you know, say in the early 1990s there was lots of debate over what the NII should be, okay. And what it should look like and what technology should it be? I think the Internet has answered that for us. NII will be IP, that is the middle protocol in the Internet stack, the Internet IP protocol, over networks. And the key to really unleash the potential is having a broad band Internet, having broad band access on and off the Internet. Why is this so key? Basically, we're in an environment where we've got a tremendous flux in services. What are going to be the real services people will pay for and drive this? And I think the answer is nobody knows and they're being invented faster than we think. You know, is it going to be just doing E-mail? Clearly not. Is it going to be down loading video? Possibly. Is the network going to be used in ways that we don't predict? Absolutely. One of the early cable modems trials got into trouble because they assumed a high asymmetry downstream which is the way I've been thinking. Most of the data will come out of content providers, downstream. So in this particular trial, there was an MIT professor who had a young kid at home. And at work he wanted to be able to see his kid. So he put his videocamera in his home, started pumping video upstream. Totally blew the network out of the water. So we'll invent all sorts of ways to do new services and it's a very unpredictable environment. When I started in the telephone network, we worried over whether the voice network traffic was going to be 3.7 or 4.1 percent. It made a tremendous difference to the engineering economics of the Bell system. Now network providers out there are doubling their network every six months. A little bit different ballgame. At the same time, technology's in flux. We used to worry about how far do you have to -- can you drive copper pairs before you had to put in repeaters? Nowadays, do you go with ADSL, XDSL, VHDSL, hybrid fiber coax, fiber to the home, which is clearly a long term answer, but how soon? Switch digital video? So there are a multitude of technology choices out there. How do you couple this technology with the services without bogging everything down? And that's where the role of IP comes into the network. In effect, in this environment there's obviously tremendous amount of risk. Risk if you're the network provider in putting the wrong capital in and seeing that stranded very quickly. Risk in the service provider if you assume certain network services are going to be available. Certain capabilities and performance, if it's not there, then you've developed your services and nobody can get decent access. So let me real quick high level toil on network architectures. What we did in the past and in effect learning from some of our mistakes in the past. Voice network. We had a vertically integrated network. Services were inclined with all the technology in the network. We wanted to introduce a new service. We went into large switching machines and developed new code for voice services. Pat White, who's one of our speakers, was instrumental in developing some of that in the SS's. Tremendous software development, long development time to get new services out there. But a lot of them are doing, you know, are tremendous service winners today such as call waiting, call forwarding, et cetera. But the network, the services, everything is tightly entwined. Furthermore, it was analog. You know, traditional voices analog, it was okay to have an analog network. The next step we went to was ISDN. ISDN was digital and is faster. So that's an improvement. But ISDN still had service specific technology. You had circuit switching. You had packet switching. And you design circuit switches around the underlying technology. And that worked if the framework was a monopoly environment where the services would be provided by the monopoly carrier. And ISDN came out of the standards environment, driven largely by the view circa 1980 and the European. Next we moved to ATM. We got closer to getting it right, but still not -- and let me make it clear. ISDN I think has a real role, particularly over the next five to ten years. Because as you'll see later, getting a broad band mass network out there quickly is a tremendous challenge. And today almost all Internet services are much better over -- well, they're all better over ISDN than over POTS modem. And actually at ISDN speeds of 150 kilobits per second. That will be adequate for most of the services people envision over the next five years. Apparently that's the view when we talk to people like Microsoft and others. so I think ISDN does have a very important interim role. ATM, which also came out of the telephone network world, now has a single underlying transport packets and cells and all that allow you to do a variety of services over top of ATM. But it still requires, before you can realize it, changing all the network out to ATM. Tremendous capital impediment to doing that and it ties still the service evolution through something called bearer services through the underlying technology. Meanwhile, the computer world developed the notion of IP and a separation of the service layer through IP in the middle from the underlying network layer. And that's been one of the beauties of Internet is this simple protocol set in the middle. People had developed all of these services on top of that protocol and they can work over any underlying technology. The only question is performance as we'll see a little bit later. But we've separated the service evolution, the competitive service market, from the underlying technology and that's why I think IP and internets and intranets will be the broad band network. So then the real question is how do we provide the best possible access to that competitive world of IP through the competitive alternatives of access technology? The result, adding new networks and services is low cost and straight forward. At least adding new services is relatively low cost. One can argue about the low cost networks. For corporate America, for high speed lines, it was fairly straight forward. For the mass market where you have literally 100 million access lines out there, you still have a challenge. But you decouple that evolution. Another important corollary which will drive the services, if you have services tied to a particular network provider, those services will never reach the same market mass that Internet. Internet will always reach more customers worldwide than any other network because you just have this narrow, well-defined protocol in the middle. So if you're a service provide, you want to get your services on the Internet to get market reach. That, of course, means it will attract more service providers, leading to more rapid service evolution and therefore more pull for more and more network demand. And what's going to happen on the Internet? In addition to all the services we're used to using today, from E-mail to any time I need to know about a corporation I go into their web page, see what the corporate information's all about, to images, all of these new services, games, et cetera, are coming. But I think there are going to be two important additional services that says Internet IP really will become sort of the universal service of the future. One is voice and video. By 1998, I think you will start to see substantive voice and video on the Internet. It's not saying, I'm definitely not saying it's going to replace the telephone network. But you'll start seeing these services in a very real way. Let me give you an example of one of the things we and other companies are doing today. Voice mail and fax. They're not real time two-way services. You're not trying to talk and tell time when you're doing voice mail. You're just listening, sending your message back. Fax is obviously not a two-way communication. Although sometimes they fly back and forth fast enough you think they are. Okay. What's happening today? People take a local server, dial up a local call through the local network where basically there are no charges to minimal charges, to a server. That server then takes the fax or takes the voice mail, okay. Packetizes it. Puts it over the Internet, POTS it out on a server say in Hong Kong or Taiwan if you started in New York City. Then goes back through, dials up at the other end through a modem and to the fax machine that set the destination. And now service is available to take care of all the intelligence. So that to the user it looks like a normal fax call. You just dial the fax number and send it. But what have you done? You've avoided all triple -- all long distance charges and all IDDD, all overseas charges. A tremendous savings, you know. Well over the order of magnitude savings. And to give you an idea of the impact of this, over half the traffic today that goes overseas on the telephone network is actually fax. That traffic is all set to leave the current phone network and go on the Internet. The same thing for voice mail. And I do know people who are even doing voice over the Internet today to overseas locations. They're willing to put up with the lower quality simply because of the lower cost. And particularly over corporate Internets where you can have high speed back bones and you control the performance. So what's our network of the future will look like. Basically, I'm proposing a network that says you need a broad band access network from the home. That's supposed to be a fancy looking home with I guess they even have the coffee pot into the network into the home. I think that's a little fanciful. But we've got people at Bellcore who could get into that. Your TV, your fax machine, et cetera. We've got somebody at Bellcore who sits in bed at night with his PC typing. So who knows? Techies. Over the broad band network which then gives you access either to today's voice network, which will still have a very vibrant roll for a long time to come, or out over the Internet. And this will become the universal Internet. Now, what are the obstacles to getting there? At the risk of offending many very good engineers who are trying to solve difficult technical problems, I think the critical problem is the mass market access. The Ciscos, the three coms, et cetera, of the world will make their routers and switchers get faster and faster. In fact, a Cisco router circa 1997 will have ten times the band width of the largest voice switches in the networks today. So the backbone switching technology will get faster. The backbone fiber optic technology we and others are working on will give us a 20 to 100-fold increase in that very shortly through wavelength multiplexing. We'll solve the traffic congestion problems. They are challenging problems, but those are being solved. The competitive marketplace will solve those. The real problem is how do you solve the mass market access? And that's what we'll spend the rest of the time looking at. What are the options today? Today you've got the Internet backbone network, Internet access points and ISPs. How do you get on and off the highway? Well, if you're a corporate user or if you don't mind paying $200 a month, you can get a private line from your home or your office into the network. And the private lines are getting faster, the price performance is improving all the time. So again, no problem here as long as you're willing to pay a fair amount of money. But if you're not, what are the alternatives? POTS? You know, 28.8 bit per second, limited to 64 kilobit, but in many applications those won't really be anywhere close to that. ISDN which can get you up to 128 kilobits per second. They both have the advantage that they work over today's copper network just by putting something into the copper loop. Then there's a whole family of what are called digital subscriber lines, so-called XDSL, ADSL, VH or Very High speed digital subscriber lines, et cetera. And I'll talk more about those. The other alternative's going through the cable network. So those are the there alternatives and then one I will talk about but don't show on that viewgraph is wireless. Now, to give you an idea what's that like and why would this be better, I'm going to show a very short tape. Elliot asked for demonstrations. Well, demonstrating access technology is sort of like demonstrating a microprocessor. Nobody cares what happens inside the microprocessor. The applications are where the action is. So this is just going to show a very simple application downloading over four different technologies so that you can see the speed advantages and why it becomes attractive. If you could role the first tape. The first one, and I'll explain what the -- the others are going to go pretty quickly. This is downloading an image over the POTS network today. So if you want to download a busy image today, this is what you'd see. And what you'll see is, yeah, you might be willing to do it, but it's not the world's most interesting thing to sit there and watch this slowly download. Now, in order to kill time, I could tell bad jokes but Elliot told me he'd break my kneecaps. So what you're going to see next is the same image downloaded over ISDN and you'll see that that's quite acceptable. Then you'll see it downloaded over a ADSL link and finally over a cable modem. And what you're seeing at the bottom is the number of seconds it takes to do the download. AUDIENCE QUESTION: Do you know the file size on that? MR. NEWMAN: See me afterwards, I'm blanking in real time on this image. But it would be a typical fairly busy photograph. I could figure it out, but I'll do it offline because I didn't memorize that data. AUDIENCE QUESTION: Modem speed? MR. NEWMAN: It's a 28.8. The question was the file size. I said I'd have to address that offline. And he asked the modem speed. That was a 28.8 modem functioning at 28.8. And this should be the ISDN. And you'll see that's not bad. I sit there and wait for that on the PC. In fact, I'm a horse person. Actually, there are horse people now putting up pictures of their horses that we download. Now what you'll see is this becomes real acceptable. In fact, it takes longer to put up local graphics than to download the image. Boom, okay? Seven-tenths of a second. That's not bad. And you'll see in this particular example cable modem will actually beat that. So don't wink or you'll miss it. Boom, okay. That's why we won't broad band access. It will open up a whole new field of applications. Hopefully, we killed the tape at this point. Yeah, good. Okay. Let's examine these alternatives. First, let me state the customer requirements. Well, we think the customer requirements are -- and we get a lot of debate on this first one, but ideally we think users want a flat rate because you don't want to really worry about a meter ticking while you're doing applications. Now, that has some problems from the service provider side, but if you ask users, flat rates awfully nice. User friendly. It should be customer installable. I'm going to say a little bit more about that because we have a real issue. Competitive corporate America has to tackle how do we get these signals around the home? I think it should be as easy to use as the telephone network. I recently had to convert at work from a Mackintosh to a Microsoft environment and everything is not yet user friendly. I think it should be as easy to use as the telephone network. It should be non-disruptive. Right now, last night I was visiting with Elliot and I went to use the telephone downstairs not realizing his wife was on the modem upstairs and guess what? Her connection was blown away. So it should be non-disruptive to the TV and telephone users in the home. And finally, ideally what we want is an online throughput. I believe 128 kilobits today would be a tremendous step forward and that's what ISDN gives us. But ultimately we'd need megabits per second. Said another way, basically the cable industry's figured out the right price structure. The telephone industry's figured out the quality of service and the computer industry's figured out the performance. How do we put them together and still make money? Let me start with a customer environment. This is not a slam dunk, getting digital signals around the home. Analog access has some real advantages. It's much more forgiving than digital. We have done a lot of studies, cable line did some studies and say getting digital signals around a coaxial network in somebody's home is hard. It's bad news. We did some more detailed studies at Bellcore and found out they were far too optimistic. Some of the problems you get into is when people put in these coaxial networks in the home, they frequently staple right through the coax. It doesn't bother analog TV. It does wonders for digital signals. They leave things unterminated like ports here. One of the early cable modem trials, the network kept going down about 7:30 a.m. every morning. I mean, everybody had noise and unacceptable performance. It was very regular. What happened? There's a woman who's very regular in her habits and every morning about 7:30 she used her hair dryer right next to one of these things. And her hair dryer spewed electromagnetic noise all over the network. So there are lots of issues in the home. And in effect you need a home network. Now, for the average Bellcore engineer, no problem. They go in and put a ten base T network in their home. I don't think my parents are going to do that. So we really need to solve the home networking problem because otherwise we're going to be stuck with using that telephone wiring and 28.8 modem. I think this is a real opportunity out here for corporate America also, but it's a tricky environment. Just dropping this inside the current coaxial network environment of trying to go to high speed, not that easy. And for those who are interested, Bellcore and others have fairly detailed studies of that environment. Okay. So what do we do? Well, there are three solutions out there. First is to use the twisted pair that's out there today. Now, that has an advantage that architecturally it's modeled after in the data communications network what's known as twisted pair allocations routing out of the wiring closet. That's been a nice architecture because it's gotten rid of a lot of the suggestions of shared bus architectures. So it looks good and once it's in deployment, it works quite well. The problem is all the deployment problems. You're no longer just inside a building. You're going out over the telephone network. What happens when the customer is more than say, well, we're talking kilofeet which is a weird measure, but ten kilofeet from the central office. You may not be able to get the signal out to that customer without a lot of additional expense. Some of the early ISDNs have gotten into trouble because they filed the universal tariff assuming a certain distribution of people because it turns out once you get out about 10,000 feet from the central office it costs three times as much to provide INSD as you are close in. They assumed most of the people would be distributed in the same population as the telephone network users today. It turned out no. The people who really wanted ISDN lived far away from the central office and want to telecommute are the people who can live up in the canyons in California, et cetera. So their users were all far away from the telephone office. Capital expense far greater than anticipated. So that's one thing. The other thing is we put this high speed digital technology out there. It starts interfering with other technologies in the loop. So there are a lot of performance problems to be solved to get this out there economically and that's a lecture in itself. It does get the performance out there and if the economics can justify it, it's one way to the future. When I first used this viewgraph, probably a lot of the cable industry would have disagreed. But Don Malone I think has even now been quoted publicly as saying in effect these were cable modems. I think the hype got out in front of the reality. And you'll see that in a demo a little bit later. Again, there is constant problems here. In the ideal environment, this technology works great. But now you're putting in a shared bus environment in a public network and there are lots of performance problems to be solved. Furthermore, this only works when you have a very high quality cable network out there. And it has to be two-way capable. Less than ten percent of the cable networks in the U.S. today fall into that category. In this category, you get fiber out to the coax and then you use your coax as your last distribution. The problem again is you get a lot of noise in this type of shared environment. So you have to have a very tightly designed network without a lot of noise. You don't know how much upstream traffic you'll have so engineering network is tricky. And so again it's one of these that follows the rule of thumb. Engineers are always too optimistic about what will happen in the next two years and that definitely happened here and I'd be happy to talk offline with a lot of the difficulties that need to be solved for this to be a reality. On the other hand, I think it's one of these technologies where engineers always underestimate the impact of their technology within a decade. So it's getting there, but lots of problems to be solved. Now, how about wireless? Well, I think we're going to see some very interesting wireless solutions evolve in the marketplace, particularly with all the spectrum being granted. There are going to be some entrepreneurial people who I believe will figure out how to use that spectrum for very effective international access. Maybe at 128 kilobits, not at the three megabits or something we want in the near future. There are three megabyte solutions, but they have a lot of limitations and economic -- are much more expensive. But let's look at some of the practicalities. And what I've shown here is an MMDS type environment. Although some of the same lessons apply to a satellite environment. In this case, the digital signal goes out over the air, into the home. It has to be received in the home, processed, sent to the PC and come back out over the telephone network. One advantage of this -- one of the problems of this is you're still tying up the telephone network. So you haven't solved the problem of congestion on a telephone network that was designed for five minute holding times and people like myself go home in the evening, put it on and leave it on for three hours, you know, I got down, eat dinner, do everything else. But why disconnect and reconnect the phone line? It's a free local call. Sorry for my friends in Bell Atlantic, but we all do it, right? But let me talk about some of the radio problems that are subtle here. This is a digital signal. Digital signals are fragile. What we found, the FCC in the same lessons that they learned when they looked at advance TV, ATV, apply in this environment. The person on one side of the street gets a beautiful digital signal. The person on the other side of the street, no signal. So what percent coverage can you get? It makes a big difference if you're a service provider trying to provide this service to know whether you can cover 50 percent of the homes, 80 percent of the homes or 100 percent. And can you predict ahead of time? And do you have to send somebody out and do measurements before you can serve them? It makes a big difference in the economics. This one's even trickier than that though. This is a digital signal and you get interference. Therefore, you may have to build a certain amount of what's called adaptive equalization into this. But suppose the customer's sitting there getting a beautiful signal. Now, clearly since this is the line of sight, somebody builds a tower across the street from them, they lose their signal. It's more subtle than that. Somebody builds a tower offline, not in their line of sight, but offline so that they get reflections. Now their service goes away. Okay. How do you predict those type of things? So here you have a happy customer. A year later somebody constructs something. They're not even aware of it. And they lose their signal. Then you have to go out and reposition their dish or whatever. So again, while this is not an easy environment to work in, I do think there will be some good solutions and we and others are working on that. I think it's really the wild cared that may open up Internet access. Let me just show you this in more real terms. We have at Bellcore a test bed in which we can take the same applications, streaming video, live media, et cetera, put them through an Internet with routers, catchers and all that and use POTS, ADSL cable modems. You know, this is a working network with all these different technologies employed. So we can deploy the same application over different technologies, put network load on it and see how it performs. What we'll now show you is ADSL downloading a video clip. So now you download the five minute videoclip of your child or grandchild or whatever to the grandparents. We'll do that over ADSL and cable modems. We're going to do it twice. One is in a downloaded network and the other is a loaded network. Could we roll the last tape? Okay. So what you're seeing is a video clip being downloaded and the cable modem's already done in eight seconds and ADSL's going to take about twice that long. Both those are pretty good. Waiting 17 seconds to then watch a video clip on your PC's pretty good. Now let's do the same thing. And now we've put a fair amount of load, but we think a realistic load, on the network. And this shows why the network engineering is going to be tricky. If we could fast forward this up to just before the ADSL and I'll tell you when to go back to -- okay, what you're seeing now is under a reasonable network load, that's now gone to over a minute. Okay. Slow it down for a bit. Okay. I think it's just -- is it just under two minutes? Yeah. What you've seen now is in this particular load because of the shared nature of the cable modem in this example, the cable one is only halfway there. So what went from under ten seconds is now over two minutes. Why don't you go ahead and speed it up to the end? I think it ends up around four minutes. Okay. Real quick. Where will the prices come out on this technology? This is the price per customer for capital costs assuming the cable network and the ADSL network are already in place and you already have an MMDS network in place to provide video service. So this is the incremental capital cost. Is this accurate? I don't know. If I had to make a pick today, this is my best guess in reasonable volume deployment. And one of the implications here is all of these are greater than $500 per customer. So if we're trying to solve, you know, serve the mass market, a non-trivial capital deployment problem, we multiply that by 100 million customers. And as I said, these prices are best picked. They'll change over time. But that is I think is reasonably optimistic when you're down the growth curve. Everybody's taking that down. I'll deny I ever said that later if you take it down. Okay. I promised Elliot and Bob to keep it down to about a half hour. So let me just for those who want I'll talk with you offline. But basically, I put up the user requirements earlier. And what we see is none of these really meet the ideal user requirements. So we've got to face up to what are the compromises? What are the price performance? In fact, the ultimate answer is going to be fiber to the home. And I believe within five years that will be the economic choice. But it's not here yet technically and now you're talking about totally rewiring America with fiber which Corning and companies like that would love. But that's again a large capital expense. On the other hand, I think what you can't do is sit on the sidelines. It's clear to me that because of what's happening in the Internet, it will increase the capital outlay if you're in the network business. It will increase your expense to serve these customers if you're a network provider. And if you do it -- and it will change customer perceptions, for better or for worse. If customers can't make telephone calls because of Internet congestion, that's going to change the customer's perception. The same thing's true in the cable industry. Ultimately, if you do it right, I think it will be a significant increase in carrier revenue. What should we do about it? Develop service independent loop technology so the services don't obsolete the network, drive the industry towards IP dial tone. Start deploying these new services towards getting experience because any new service introduction takes at least a decade. But now's the time to start getting experience. And finally, I think what we're going to see is the surfing the net worldwide we'll probably see 120 million users. It's a terrific market to reach. And what you see is that's now become a tidal wave. Thank you, very much. [applause] MR. PEPPER: Thank you, Stagg. What we're going to do is we're not going to take a break other than to have somebody come move the podium. And we're going to invite our panelists up. A couple of things of a housekeeping nature. One, since we don't really have time for direction questions of Stagg, we're going to ask Stagg to sit up here with us. I imagine a number of the panelists including David Reed from Cable Labs might have some comments on some of the things that Stagg said and we can involve Stagg in the discussion. Second, please because we're transcribing this, please use microphones for asking your questions so that we can actually get that over the net as well as transcribe it. And then also we would be putting up on our web site, the FCC's web site, the transcript of today's proceeding. Plus, if Stagg promises to get us the graphics electronically, we'll put all of the viewgraphs up on the web site as well. So if I could invite our panelists up and see whether we're going to have somebody -- I don't think we have any other viewgraphs. If we could get rid of the viewgraph machine. Our first panel, Jeff is the mike on? I think it is. Our first panel we've asked to focus on competing models for higher band width. We're very, very lucky to have the group that we have with us. Les Vadasz is Senior Vice President at Intel. For those of you who follow Intel on the computer industry know how important Les is to the entire computer industry because of his importance at Intel. He's one of the real deep thinkers there. It's always a pleasure to see you. Pat White is Vice President for Research and Development at Bell Atlantic and prior to his current assignment he was VP in the Strategic Planning Organization where he had responsibility for telcom strategy. Pat is an engineer, has his Ph.D. in engineering from Northwestern. Pat has been in our early discussions and inquiries into this issue has been extremely helpful to us in understanding the local exchange network architectures and how they are having to adapt to data services. John Curran is the Chief Technology Officer at BBN. Of course, BBN is one of the parents of the Internet. Doug Morgan is Vice President of Marketing for Winstar Communications. And Doug we asked to come specifically to talk about some of the wireless issues that Stagg raised. and Winstar, of course, is really developing a brand new, a business for providing wireless access to networks. And, of course, we see this as being I think maybe Stagg we wouldn't call it a wild card. We see this as really one of the exciting alternatives to solve some of these band width problems or questions. And then finally, David Reed who was here for a number of years as a highly valued colleague. And we were extremely sad to see Cable Labs lure him away. But I think it was more the environment of Boulder, Colorado that lured him away. And this was actually a very wise choice. Anybody who is faced with the decision of living in Washington, D.C. and living in Boulder, Colorado and didn't choose Boulder, Colorado, we'd have to question his sanity. But David I think is very sane. And by the way, David was critical in many of our decisions in the PCS proceedings here developing what was really the only technical model of demand on band width and on market projections and he turned out to be very, very accurate. So we always rely very heavily on David's advice on issues. There's no particular order here. We've asked each of the panelists to start off with two or three minutes so that we can get this discussion going. So why don't we just start with Les and move right down the table. Les. MR. VADASZ: I guess this is on, is it? Okay. I'd like to thank you for the opportunity to participate in this broad band forum. I think this is a very timely event and I'm glad that I have a chance to address this issue with you. Intel is a major supplier of microprocessor components and systems and particularly to the personal computer industry. And we are very familiar with the needs of this user community and this will be the basis for my remarks this morning. Let me begin with the obvious. A consumer wants to use their PCs to communicate. I think that the enormous growth that we have seen in the Internet access and the fact that electronic mail has now become an integral part of our everyday life is a very good indicator of this trend. And if you look at how consumers buy personal computers, virtually all of the personal computers that go to the home come equipped with a modem these days. We are living in the information age and the consumers participate in that by using their PCs. Millions and millions of homes are online already. Unfortunately, despite the government's call to build a national information infrastructure and despite the Telecommunication Act of 1996, the personal computer user is more of an after thought in all that happens when it comes to deploying network capabilities. This is a rare opportunity to address some of those issues. What is our problem in effect is there is no meaningful PC communications service available. And that's kind of obvious. The consumer can purchase many services. He can purchase electricity services, gas, telephone, TV, and in fact the FCC has addressed many of these consumer demands over time. But we cannot buy a reasonable PC communication service. We merely piggyback on the telephone service and get a very inadequate connection. I think that we have heard earlier that there are many technologies available to solve some of that problem and certainly an Internet connection through a telephone line and a modem is not one of the adequate solutions for the future. This is happening in spite of the fact that nearly 40 percent of the U.S. homes -- that's nearly 40 million homes -- have personal computers. So any numbers that you add up in this arena is going to be big financial impact to all considered. So what are these basic communication services? I think it's important that we characterize this and band width is only one of the elements. I think Dr. Newman did an excellent analysis of what are the various characteristics and I believe I will be redundant in some of my discussions of those, but he characterized them well of what the consumer needs. First of all, high band width. There are many technology choices. Virtually every service provider has an opportunity to upgrade their networks with new technologies to provide more meaningful PC communication services. Second, instant access. It is unacceptable to wait two minutes to connect to a network. You wouldn't accept that from your TV. You wouldn't accept that from your local area network at work. There is absolutely no reason that you should accept that in your home environment. Third, plug and play service. It ought to be easy to just take your computer and provision a service. We are not there. Most of you, many of you who have tried to connect through services like ISDN I'm sure know the problems that you go through in trying to provision a service for your PC. Fourth, multimedia capability. Voice, video and data communication ought to be seamlessly integrated into the service. Fifth, store forward capabilities. Voice mail, electronic mail, fax and other store forward capabilities ought to be service offerings. Sixth, security. Security options ought to be available because both electronic commerce and workplace connection requires protection from intrusion. Seventh, affordable pricing. What's affordable pricing? Well, telephone service is affordable. We know that. TV service is affordable. We know that. I believe that a PC communication service has to be in the same order of magnitude as those services available. Now, with these characteristics in mind, a meaningful communication service need to offer the following. First Internet access. As Dr. Newman said, this is baseline. This is the digital dial tone. Second, multicast broadcast services. News and other information services ought to reach the user when they happen. There are many technologies occurring, happening, that allow you to deliver these kind of capabilities. And third, our connection to our workplace through a modem is inadequate. We need these capabilities and we need telecommunication services offered through this infrastructure. Now, volumes have been written about the economic impact of the Internet and all of these in my opinion depend on deploying a meaningful baseline PC communication service. And the economic potential of this service in my opinion is enormous. If you look at the 40 percent of our homes, 40 million homes with personal computers, a baseline service should represent in excess of $10 billion revenue. On top of that, if you think about the investment that both the consumer will make and the equipment suppliers, the service providers will make, there is a significant equipment business associated with this as well. And we know that the consumer will pay for value. The current investment rate, current demand for PCs for the home and rapid increase, rapid demand for satellite TV services is indication that the consumer will pay for quality capabilities. Now, I do not intend this to be a call for a new set of regulatory regimes. And in fact, I don't even intend this to be a call for some technical specifications to deploy the services. I believe that it is the industry's job to do that. However, I would hope that the Commission when it faces the issues that it struggles with looks at these capabilities, looks at this PC communication service needs and makes those decisions with those issues in mind. How will our decision impact the development of these PC communication services? What will the service providers do as a result of our actions? How will a competitive market be developed where different service providers using different technology means are going to compete for the dollars of the consumers. Those are critical decisions that I believe are going to impact how the future national information infrastructure will develop and how the economic benefit from all this technology will develop for all of us. Thank you, very much. [applause] MR. PEPPER: Thank you. If we could also try to keep these statements to two to three minutes. Otherwise, we're going to run out of time for any discussion. Thanks. John. MR. CURRAN: Good morning. I'm John Curran. I'll keep my comments brief. This will be facilitated by the fact that I'm from Boston and I speak rather quickly. If you need to catch up, there will be transcripts I'm sure somewhere. Basically, my job would be to live between 12 and 24 months out in where the Internet is today. I used to try to live three years out, five years out, seven years out. Now if I can keep 12 months in advance, planning and where we're going I consider myself lucky. The fact of the matter is that the Internet is one of the most remarkable changes that's come across telecommunications to date. And I want to highlight how the Internet works for this forum to describe some of the interesting cultural differences that are happening and things that we need to keep in mind when we try to look at the issues of access reform and how we're going to bring about broad band infrastructure for everyone to use. In the U.S. alone, we have over 3,000 Internet service providers. You know of probably, oh, the top ten of them, the people who run some of the national backbones. But you're probably not aware of the other 2,900 that are out there. Some of them are in every community. One of your next door neighbors almost certainly is an Internet service provider. The fact of the matter is that the Internet service industry has grown at such a dramatic rate and had such remarkable success not because of the degree of its coordination at the end, the amount of architecture that we've put in place, but in fact the exact opposite. The Internet is an example of success in decentralization. The fact of the matter is that in the Internet we work in a very simple basis where we standardize and we work on cooperation on the bare minimum facilities necessary to keep things communicating and leave the rest to an open market model. Just to give you an example of this, over the last three years, the Internet has gone through three different backbone models. We went through a model that we had one single nationwide backbone. We went through a phase in time where we had several nationwide backbones of large IP routed networks that were interconnected through a few key locations and most recently we've actually gone to a model where the major backbone providers, providers such as some of the inter exchange providers BBN and others have built a model where our backbones are interconnected in multiple locations. Those are three dramatic changes that happened in 36 months. I can't emphasize more strongly the fact that the Internet has to be able to evolve rapidly and in fact what we're really looking for going forward for the ability to continue to evolve the Internet is to create the same market conditions in areas such as local access that have existed in the abilities to purchase wide span high speed broad band communications. If we have that open market, the Internet service community, and I mean thousands of entrepreneurial businesses, will explore every backbone model possible, will explore every access architecture. We will see a rich tapestry of these coming forth. The fact of the matter is that the Internet service provider industry is using the infrastructure as any other business. Today there are corporations out there that have hundreds of dial up lines and there are bulleting board servers and others who make use of local access circuits and local switching infrastructure. The Internet service providers have done the same thing. They are looking at the rates that have been provided and they're trying to do what makes sense. I don't think there's an ISP out there that doesn't want to pay their fair share of the cost involved in servicing that infrastructure. It is not a case that there's an industry looking per se for a free ride. What you have is you have a case of certain pricing in place and people trying to do the best business decisions to deploy it. Now, clearly there are some hard issues here. We have a situation where all of these pricing, all the pricing for local access has been setup with certain assumptions such as loading rates and busy rates that no longer apply because of the popularity of the Internet. To this end the fact of the matter is we need to fix some of those basic underlying economics. We need to make it such that the price of a local access loop and the price, for example, of going through the switching infrastructure is clear to the industry. So that they can make intelligent decisions about some of the technologies we saw presented earlier today. I can't think of a better way to decimate an industry than to establish an artificial form of cost recovery for these functions. The fact of the matter is that the FCC in its wisdom has already undertaken open interconnection activities which will over time provide a way for this robust industry to look at the complete components that it's using and make sure it's paying the adequate cost. And if indeed there are areas where the current models don't make sense in a different architecture for surfers, whether it's different service architecture or different technology is necessary, these will be made naturally by a robust marketplace. I asked that this forum and the Commission in general take these comments into effect when going forward. Because what we have here is a remarkable success which under the appropriate forms can be turned into a remarkable opportunity to deploy new technologies and to harness the incredible dynamics of the information in advancing technology and communications to the mass market. If we should make a mistake of trying to architect a particular reform or a particular solution on behalf of the entire industry, I think we can safely predict that the robust entrepreneurial vision of the Internet won't be able to be harnessed and in fact we'll lose that energy going forward. That concludes my remarks. Thank you. [applause] MR. PEPPER: Thank you. MR. WHITE: Okay. I guess it's my turn. I'm Pat White from Bell Atlantic. Is the microphone on? Thank you. I know that from the previous comments from Stagg Newman and others on this program so far that the rest of the comments have tended to focus on the access to the network being the focus for most of the problems. Well, I'm in Bell Atlantic. I have access to a ten megabyte LAN within the company premises. And this LAN is connected to the outside world over a T1 or T3 connection. And I can say when I'm downloading the files I've never ever seen my PC record anything approaching even a megabyte per second. It's usually a few kilobits per second at most that you get. So clearly, no matter what we do with access, we're still going to have a significant problem accessing the Internet. And in fact, if you start to carve things up a bit you'll find that a lot of problems exist in the firewalls in the premises. A lot of problems also occur within the PC. Because as my friend at Intel keeps increasing the processor speed, Bill Gates keeps using it up with more and more bloat ware. So the amount available to actually handle all of this processing is very, very small. We also have congestion on the Internet because with the kind of structure that the Internet has, with everyone engineering their facilities according to their own rules, there's no end to end performance or end to end view of the Internet. We find that there are a lot of times when you could experience packet loss going in excess of 10 to 20 percent on some occasions in the Internet. And finally, there are the servers on the Internet. They're also heavily congested too. So what I'd like to offer here is that instead of focusing excessively on just one aspect of the Net we begin to think about it more expansively, more globally. There are a lot of things that have to come together before we can see the performance capabilities that Stagg Newman talked about. I couldn't agree more with my colleague at Intel on the needs for the PC. I agree completely with him. We need Internet access. We need multicast and broadcast capability. We need some kind of direct connection. And furthermore, I'd want more. I think you also need to have continuous connection available at all times. The only problem I think right now, in fact we know how to solve a lot of these problems, both the cable industry, the phone industry, wireless companies and others, is the real question in my mind is who's going to pay for it. Clearly, there are forces who believe that the vast majority of telephone customers who don't use or need these capabilities at least at the present time should all pay more than their fair share. And I think that we really have to examine that carefully. It seems to me that as a minimum the cost for providing these services should be more focused on the people who are actually using, accessing the Web. We might have 40 million people with PCs in this country, but if you look at Internet access or online access you find it's less than 15 percent of households in the U.S. today who ever access the Web. So should the rest of us who don't need or use these capabilities have to finance the 15 percent who need to access it? That's the end of my remarks. MR. MORGAN: Good morning. I agree with Pat too. You have to take a look at this thing from a total network perspective. And on Stagg's comment about the wireless being the wild card, very frankly, I'm just glad that the industry has moved. You know, a year and a half ago if I was sitting here in this chair, you wouldn't have heard the word wireless. And now with a company like Winstar that can serve 150 cities around the U.S., possibility of a million circuits of DS1s, people are looking at us and saying, these are a real part of the network. So I'm glad to be mentioned. You can kind of tell me whatever I want. Just keep mentioning us a whole lot more. We're really happy about that. Let me say a little bit about what Winstar does and we are the company that really has the largest experience base as far as providing broad band access using wireless technology. Our businesses that we provide services and technologies that really complement the local carriers by providing them local access communications using a bit of different technology. that being broad band radio communications. We also take maybe a bit of a different view than some of the carriers. and I've had experience on both sides of the house. Mr. Hundt in his comments this morning talked about the last mile. A lot of people talk about the last mile. My view is if the consumers are really going to come to the fore and get what they need to move their businesses and their homes forward, we need to think about this as the first mile. Take a different view of the network. How do you get that first mile connection into this wealth of information that's out in the backbone? Let's take a little bit different view of this thing. As I said, we use a bit of a different technology here. We're using 38 gigahertz spectrum with digital broad band radios to get the connectivity. And from the standpoint of where we are today, Winstar is operational today providing service really to a variety of people in the industry from caps like ACSI and ELI and Brooks Fiber and the range of competitive access providers as well as long distance carriers like MCI, Arbucks. I think that we have Pacific Bell on the panel in the next session is a client of ours also. So we provide a number of services to the carrier marketplace today. We do know how this stuff works. We know how to go in and we do know how to make it reliable and I'm sure that we'll get some questions on that. I had mentioned capacity. There is a tremendous amount of capacity available within this 38 gig spectrum that we have. We've done calculations within our major markets that we can provide well over a million DS1s of capacity over the next couple of years within our major markets alone. So there's a huge amount of capacity available. I also agree that you have to take a look at this at a whole network view. The intrigue of wireless is really quite simple. It's an incredibly fast time to market. the capital gets optimized because I'm not digging up streets and houses and everything else. So you don't have any of the construction costs. I never have stranded plant. If bodies and people and buildings and businesses move around, I don't have stranded plant to worry about. So all of those benefits are a tremendously appealing case for wireless. You still have to take an entire view of how you role a product out to the constituency to make sure that you can keep the five nines of reliability that we built in this network that runs this country. It means you have to understand network engineering, how you provision your services, how you have your logistics in place to spare all of these and everything else. So what I'll say for all of our folks that are in the technology world, and I've had the benefit of working with Intel and actually quite a bit with Cable Labs in a prior life. The technology here, we have the greatest technologists in the world that sit in this country and many people sitting in this room today. The technology isn't the issue. The issue is how do you deploy this and keep a network that runs end to end? How do you provision it? How do you maintain it? And how do you manage it? So what I'll say to everybody that's sitting here in the room is you can get very excited about the technology and about the promise of the technology. The hard work becomes how do you deploy it in an end-to-end, cost-effective solution to bring services to the people? We think we have an incredibly important part of the solution, but we all need to work together to bring it to the people like you that need these kind of services. Thank you. [applause] MR. REED: Well, Bob, with two or three minutes, one of the results of moving to Boulder is that I do speak a little slower I think in contrast to my earlier colleague. I wanted to comment a little bit on the presentation earlier by Stagg Newman that set the paradigm as being IP dial tone. And perhaps I'm still hungover from my video dial tone days, but I don't like the term dial tone. The paradigm about the cable industry I believe will likely be a winner here has got nothing to do with dial tone or any connotations that might be associated with that term. And what I'd like to do is to address some of the technological characteristics of the high speed data service delivered over cable systems. And specifically I want to make two points. One is that the shared access network architecture delivers several key benefits to subscribers. So from the other presentation I'm going to spin that entirely on its head. I do believe, like I said, I think the cable industry is a winner with the shared access architecture. And as a result of the benefits you get with that approach, the performance of cable based high speed data service will far exceed most Internet access service available over telephone lines. And to that we all agree from the earlier presentation. The second point is that the cable based service for high speed data is being offered today in several areas and the overall deployment that is accelerating. Those who do have it commercially available are somewhat looked upon as the privileged few now I understand. But it is being, the roll out is being accelerated by many of our member companies. The next result to date of that roll out is that the subscribers are very enthusiastic about the service. The modems work well. The system works well. And it's a positive experience. And so I think you'll continue to see that roll out. So those are two important points. Now, turning to the benefits of the shared access medium, what are they. Very briefly, what I mean by the shared access to begin with is in contrast with the dedicated switched network paradigm that you had over the telephone network today that's a single high band width pipe that's shared among users. In the context of the cable industry, you're talking anywhere from 10 megabits to 30 megabits that will be shared by multiple users. And because it is shared and there is a need to set rules to share that band width in a fair and efficient manner. And that set of rules is often referred to as a Mac layer, the Mac protocol. And the Mac stands for Medium Access Control. That's the true performance enabler here and that's what really causes the performance to be substantially better with this service. So the benefits of the shared access, one, is it's well suited for the bursting nature of today's Internet traffic. If you look at how the traffic varies if somebody's surfing the Web, they need a very high transfer rate for very short period of durations. The picture we saw earlier had burst. And to the extent to which you have the band width available to accommodate that, it can be served up very quickly. And in fact with 10 to 30 megahertz per second, the band width you have over a cable service is faster than your PCs can handle today. Now, with Gates laws we discussed earlier, that will probably change. As there's availability, it will be gobbled up. But to be sure right now it's quite a bit faster than what PCs can do today. And at Cable Labs we've modeled a 60 second surf and in fact I don't have it available with me now, but if anybody's interested give me your card or contact Cable Labs and we have a little primer on cable modem performance. I'd be happy to share that with anybody that goes through exactly how the performance of a cable modem will operate. But the traffic itself because the bursting lends itself, the good news is that it lends itself well to what's called statistical multiplexing. And so that means many more users can use and share this access pipe than just dividing the band width up on a dedicated basis to each user. So that's the first benefit. Second is that the cable's large band width permits excellent performance. In its primer, for example, up to 200 simultaneous users can jointly do this simulated surf with no degradation in performance. Now, this architecture I'm talking about is the target architecture that the next generation modems will be operating under. But that's where the cable industry is going. And under the loading scenario, if you double that to 400 simultaneous users, that takes twice as long to complete. So if this is a 60 second surf, it takes one minute with up to 200. If you move it up to 400 simultaneous users, it takes two minutes. And that's still faster than 128 kilobits per second ISDN. So clearly the performance here is quite acceptable. The third benefit is the cable network architecture delivers a scalable approach. We're not talking about two simultaneous users over Washington, D.C. We're talking about 200 simultaneous users over one fiber node. And the hybrid fiber coax architecture of the cable system. And those nodes are being designed typically now anywhere from 500 to 2,000 homes. So you can go through the numbers there. You can see that you need quite a heavy loading for you to start experiencing any degradation in performance. Fourth benefit is that subscribers can stay on line and not consume network resources. So for E-mail perspective, hopping on to get information quickly, it's no problem with the shared media approach. Another benefit is that our users are connected to the same information pipe and therefore broadcast applications can be efficiently offered. So the streaming audio and video, for example, from this conference to the extent someone wants to view that or more than one user wants to view that on a particular node, with the multicast capabilities that are being built into the Internet, you're going to be able to efficiently provide that. So that's in a small way doing some future proofing in terms of the new services. The final benefit here is that designing the architecture to the extent it's possible and no one's a perfect predictor of the future to accommodate the new services. The Mac software will be downloadable. With the HFC architecture, you have ways to scale capacity to meet the need. So the upshot here is the cable industry thinks this looks like a real market winner. Because the real benefits are tangibly better than what you can do today. Second key point quickly is that the service is rolling out today. The customers are very happy with the performance. By Cable Labs count, there's over two dozen vendors who are building, in an advanced state of building cable modems. Future direction. Cable Labs has been participating in a specification process that will help put out a standard cable modem so that we'll be in the paradigm where you'll be able to take the modem from across cable systems throughout the country and that's very far along. Vendors are commenting on that. It's already been proposed to the vendors and they're coming back to the cable industry. So that is nearly complete. Several of our member companies have the commercial services available and they're really too numerous to list at this point in time. To be sure, different business models are still under consideration and we're still developing system-wide considerations and solutions to address the band width that the last mile you do have to address the other parts of the Internet. So in summary, the two key points. There are very substantial benefits and we think this is going to be a very dominant paradigm in the future for how folks will use the Internet. And it is happening and does take time to roll out. There are a substantial number of folks who use the service today. MR. PEPPER: Jeff, if you could just leave the mike's on the tables now that would help. When we get to the questions, there will be somebody I assume with a microphone who will walk around and people can ask questions into the mike. Is that person in the room yet? Great, thanks. Maybe I'll start off with the first question. I think it was John who said, maybe it wasn't. It was somebody -- I think it was John. Technology's not the issue. Or maybe it was Pat. It was Pat. Technology is here. David said it's a matter of time. If it's a matter of time, how soon? I mean, realistically. And if it's not a technology question, is it a market question? Is it a competitive question? Are there regulatory issues? Are there things that are now in the regulatory process that are impediments to the deployment of these technologies? Because I think these are the kinds of questions that people are asking. We're certainly asking ourselves. Does anybody want -- Pat, do you want to start off with that? MR. WHITE: Okay. I'll start off with that, Bob. Within the Bell Atlantic region currently, some 98 percent of our central offices are capable of providing ISDN services. And we are indeed marketing ISDN services pretty aggressively. We also have done trials of ADSL1 which is the version of ADSL that Stagg Newman talked about earlier which is the 1 1/2 megabyte version. We are also planning to deploy ADSL2 which is a six megabyte version of ADSL. You know, those considerations are underway right now because the technology is available to support up to six megabits downstream to the customer. One of the, and I guess I should also point out that we have an SDV, Switched Digital Video trial going on as we speak in Dover township in New Jersey and there is a more up to date system being planned for Philadelphia. With our switched digital video system, we will be capable of delivering up to 52 megabits over a twisted pair, very short twisted pair distances directly into the home. So some of those things are beginning to happen. However, there is an element of risk I should point out. Because investments in the network like any other investment, maybe unlike most investments, tend to have a pretty long lead time. You know, there's a lot of money, a lot of construction to be spent. And generally, you find that consumers are more interested in a new capability when a lot of people have the same service. Communication services generally have that property. They're useless for one person. But millions of people they're more interesting. So we need to see or at least we need to have regulatory policies that are consistent with the fact that you have to basically lay out a lot of money over a long period of time before you begin to see a return on that investment. And the extent that the regulatory framework in fact we're still trying to struggle with the implementation of the Telecommunications Act to the extent that that increases the risk for new investments, then I think that could have an impact on the willingness of operators to basically aggressively deploy SDV or ADSL technology. MR. PEPPER: Les, you're shaking your head. MR. VADASZ: I have a more pessimistic view of it obviously. I think if you look at our telecommunication carriers today, they spend more significantly, orders of magnitude more effort in trying to get into each other's business rather than trying to develop new business based for PC users. And this issue of who pays, well, in real businesses, in real competitive markets, you have a right to compete and you have to earn a return. You have to put capital at risk. And what we would like to see is the communication carriers recognize the PC user as a real viable economic opportunity. I guess that's what I -- MR. WHITE: Bob, can I respond to that? MR. PEPPER: Pat. Yes, please. MR. WHITE: Bell Atlantic currently spends in the neighborhood of $2 billion a year every year on new construction and maintaining the network equipment within our territory. I would hardly say that that is compatible with not investing in the network and looking elsewhere for returns. Indeed, we spend most for capital right here within the states that we currently serve. MR. PEPPER: Doug, I was just actually going to ask Doug, wireless, you pointed out that long lead times were maybe not the same in wireless. You can have more of an incremental approach. You don't have to worry as much about stranded investment. How does wireless and your business plan? MR. MORGAN: Our paradigm is much, much different. And today we're serving business communities and multi-dwelling units. We're not at the residential level. But I certainly assume when this technology comes down we'll be able to spread that. But where the technologies play out if you look at the cost of any access provider to go into the building, today it's somewhere between $40,000 and $100,000 depending on what city you're in and how much construction you're in and everything else go into that building. And I can understand -- Bell Atlantic saying I don't want to spend $100,000 to go under a building. Our approach is much different. We put a 12 inch dish on the roof, a couple of small radios inside and I have access to the building. So we're actually a unique competitive advantage for the carriers that we work with that they can use us to go into buildings to get entry to provide high capacity services and quite frankly if the carriers find that after they're in a building that there's enough capacity that warrants putting fiber in. We're a big believer in fiber. We use fiber ourselves to connect the nodes together. We think it's a tremendous technology, just not financially feasible if you have to build it on a building and hope that it will become kind of a model. So we give people a unique advantage. They can go into a building when there's enough capacity that they want to go ahead and build fiber behind us, they can go ahead and do that and we can take our plant and move it somewhere else. So we see ourselves as a real enabler to move some of this new technology forward with a much different financial plan than anybody else has in the industry. MR. PEPPER: Now, there's a question Stagg raised about wireless and its ability and interference characteristics and potential problems which leads to the question of the network liability issues that the Chairman raised. MR. MORGAN: Sure. I'd be glad to address that. As the largest provider of these kind of services, I won't tell you that we haven't had a learning curve as we've gone through this. If you take a look at vendor literature that talks about a 38 gig range, four and five mile distances and things of that sort, it will certainly work on those kind of distances. It won't work to the level that any of you will be happy with. So we have found that the nice point about the technology in our range is it's very much a point and shoot technology. This is not difficult. If you have line of sight, we can do a distance. The real thing we've had to find out is you have to make sure that you keep your distances short which means that depending on the region of the country that you're in, somewhere between a mile and a quarter to a mile and a half to a little bit longer as you go past, if you want to keep the five nine path reliability. Physics are physics. They haven't changed in 100 years. And the real thing you have to do is make sure that you have to keep the paths right. So once we have gone through the proper types of network engineering, you'll lay out your connection points in such a manner that you keep all of your paths short. And that also reduces the reflection and other people getting in the way of it. Once you do all that, you can build an incredibly reliable robust network, but you have to understand the engineering pieces and don't believe all the things you read from the vendors. MR. PEPPER: John. MR. CURRAN: Let me pickup on the Internet reliability issue. Certainly there's a real issue here. We're in a situation because again of the wide number of providers going forward and the very diverse technologies being used to deploy this, where we have different standards of reliability. In fact, there are some pieces within the work that I'm sort of pleased have different standards. The Internet has made it into countries where 386 PCs are state of the art, where the modems are older than many of my staff. And the fact of the matter is that the Internet is unique in its ability to connect all of these communities. There are rural networks out there, for example, for educational purposes that wouldn't meet the requirements of my business customers. I have 3,000 business customers connected to the network. They have very high requirements. That doesn't mean I don't want to connect to those other networks. You have to recognize when we talk about ability that we want a core network that is very robust, but the Internet accommodates almost anything from multi gigabit per second links all the way down to 1,100 baud VLF communications. That's all on the Internet somewhere. So I think it's important -- two things to keep in mind. First that when we talk about Internet reliability, I don't think we want to talk about ubiquitous in that reliability until we're sure we're willing to pay the cost for every corner of the Internet, where the Internet today is still going to corners that are fairly rough. The second item was just the fact that the Internet technology's changing very rapidly. I routinely run into people who are waiving around 15 year depreciation schedules and 30 year depreciation schedules. And I think about the fact that we're doing doubling every year. We're doing upgrades on a 12 to 18 month basis of all our key technology. I think that there's going to be a period of time when the Internet settles down and those periods lengthen out, but it isn't today. And I think that it would be wise for the Internet to have one-tenth the number of years that the telecommunications industry has had before it has to go tackle the reliability issue right now. MR. PEPPER: But how does this resolve Les's question of the real issue is mass market? Les. MR. VADASZ: I think that reliability is very important depending on the application. For example, today electronic mail has become a mission critical application in many of our corporations. However, I think that we have time to evolve. I think something like our telephone network is probably the most reliable, most impressive reliability that I've ever experienced in my whole company. MR. WHITE: Accepted. MR. VADASZ: We will get there. We need to get there. But I don't think we have to be there today. I think we have time to evolve as we deploy more and more applications on the network. I think that where we have to get very rapidly is broad band connection to the home. And, yes, indeed there are many other issues to be solved, but again, they will be solved incrementally. But I don't want to lump that issue of broad band connectivity to the home into just one of the issues that we have to solve along the way, among the 100 issues. That is critical now, urgent, and it's not happening. MR. PEPPER: Stagg. MR. NEWMAN: I just want to echo that. I believe the competitive marketplace will solve a lot of the problems. There are certainly major engineering problems in the backbone connectivity performance, et cetera. And there are going to be, there are lots of technologies available. There will be even more. I think what's key is, one, we're allowed to use all of those. We'll need wireless to get to some homes. We'll need cable to get to some. We'll need ADSL. So you need to let all those deployment technologies, all those technologies roll. LMDS, Local MDS is one. It's 28 megahertz, wireless technology. My guys weren't very enthusiastic about it. HP and TI are pretty enthusiastic about it. Finally, we observed, my guys are in New Jersey, fairly foliage rich, rain intensive area. You know, TI's out in Texas and HP's in San Jose. And the short form is your enthusiasm for LMDS is inversely proportional to the number of leaves you see when you look out the window. So we need to allow all those technologies to be deployed. But I agree with Les. The critical issue is the mass market. That's where the tremendous cap X will have to be spent. That's where you either have to dig up streets or have access to spectrum. And frequently, it's not the technology costs. It's the digging up the streets that drives construction. And so I think that is the critical issue that we need regulatory, economic, all of our best thinking to solve. MR. BERNENGER: Dan Bernenger from Vocal Tech. I just wanted to add one other point to the reliability issue in that this time around reliability is also an application issue, that there's nothing I can do to my telephone to make the telephone network more reliable. If you give me unlimited band width and you tell me, well, I'll give you ten meg, but I lose ten percent of my packets, I can do things on the application level by sending multiple packets out of the same quality or the same content and deal with it at the other side. So reliability is now also an application issue, but it's kind of a tradeoff. If I'm paying a lot for each of my packets, then I'm not going to put redundancy in there and I'm not going to do various interpolation things or whatever. So consider that as well. The applications has contributions to the reliability. MR. PEPPER: Thank you. Are there other questions? If you could stand up and identify yourself. And if you could raise your hand in the back so the microphones can get to the next people. MS. WOODARD: My name is Gwendolyn Woodard. I'm President of Worldwide Education Consultants. The questions are for Mr. Reed and Mr. Morgan. Mr. Reed, if I am a subscriber and I subscribe to the cable network services that you have to offer, if my lights go out, will I still have access to your network? MR. REED: The issue of whether you have emergency powering typically the way the cable industry is approaching that right now is for cable television services, there are a number of our member companies -- there's no uniform strategy, but a number of uniform companies do put in backup power to handle that situation. Unfortunately, if your lights go out, the television's out. So you can't really -- there's no reason. The same thing's true with the personal computer. So there's not a lot of value in backing that power up typically. GWENDOLYN WOODARD: Okay. Mr. Morgan, you talk about wireless technology. From my understanding, if you're going to go from Point A to Point B and I send information over the wireless technology, it's still going to have to hit somebody's network before it gets to its destination, whether it's wired or wireless. MR. MORGAN: The technology that we deploy is for your access to the backbone networks. So you're correct. If you're going from one educational institution to another that's five miles away, you'll use me as the access to get onto the network and then you'll hop off at the other side. Within a campus type environment or within a private network environment, many people will use our technology right within their campus. If you don't want to tear up a campus or something of that sort, they'll use it as a private network within their school. But anything that goes off the network, they'll use us to get off the network to a public network somewhere. GWENDOLYN WOODARD: Okay. And also from my understanding, no matter what we do, we're going to still have to use the wires or wireless capabilities of the LECs and IXCs in order for you to get anywhere as far as the Internet is concerned. You're still going to have to use them. MR. MORGAN: I guess it all depends on where you take your definition there. We do have places where we'll go from a school directly to an Internet service provider. Obviously, the service provider is connected to the LECs and to the PSTN somewhere. MR. PEPPER: In the back. MR. SCHMIDT: My name's Bob Schmidt. I'm with National Digital Network and we have a PSTN test here today that we would invite all of you to see. But my question is really to Mr. Pepper and the people here on the panel from the FCC. Today we have to come to the Commission and play mother may I to do what we are capable of doing from a technology standpoint right now. And the people in the room who are users and the gentleman, Mr. Morgan has a system up and operating and I commend him. Because I think wireless has flexibility to do some things. But we've got to get the regulatory scheme caught up with the reality of the marketplace. And I think what Mr. Newman said about MMDS is true. It is another wild card. But we want to be in the game. And in order to be in the game, we need to have the regulatory environment allow people like ourselves to bring forth the capacity to deliver these services. So my question is how soon will we have that, Mr. Pepper? MR. PEPPER: Well, actually we can put this record in that proceeding as well for the MMDS flexibility. STEVE STEWART: Steve Stewart with IBM. Two questions for Pat White. And I would like David Reed to also answer the second question. And then if Stagg Newman has any comments on that, that would be useful too. First of all, Mr. White, you mentioned that Bell Atlantic is investing about $2 billion a year in networks. Can you tell us roughly what percentage of that is going to ADSL investment or other broad band networks? The second question is if you look at your ADSL and other broad band technology deployment over the next five years, can you give us a rough percentage of the number of homes that would have to be available at the end of '97, at the end of '98, et cetera. And then for David Reed, if you could offer the same percentage deployment over the next five years for cable modems. Thanks. MR. WHITE: I don't have that data off the top of my head, but I'll be glad to get back to you with it. STEVE STEWART: Okay. MR. PEPPER: Kevin Werbach wants to ask a question. Kevin. MR. WERBACH: Well, a little more than a year ago, some of us remember the 1996 Telecom Act passed. And back then everyone was talking about, gee, you are going to have these big band width networks. Cable is going to provide broad band to the home. Phone companies are going to provide broad band to the home. It's all going to be glorious. And here we are a year later and none of that seems to be a reality for the mass market we're talking about. What has held up that sort of development? Is that just international exuberance? Or are we really going to see those kind of developments taking place? Who wants to take a stab at that? MR. MORGAN: Wait until after the market closes. MR. WERBACH: Pat, do you want to take a stab at that first? MR. WHITE: I guess I get all the tough ones, huh? Well, I'd like to start commenting on that by going back to a point I made before that there are a number of uncertainties right now in the industry. We talked about the fact of the technical uncertainty that exists. We do know that a lot of the technologies we're planning to deploy work well in lab environments. When you take those equipments out into the field, you find a different thing, a different situation altogether. So we have to make adjustments for that. On top of the technical uncertainties, there's the new uncertainty with how the Telecommunications Act will be implemented. The extent to which we have to price our new technology below cost and make it available to people who are not taking an investment risk that we're taking, you know, leads to some caution that I think any prudent manager would have to take into effect before making a bet. But in spite of all of these things, Bell Atlantic has been as I said before moving aggressively with upgrading the local access. 98 percent of our central offices are ISDN capable. And in addition to that, we ran trials I think as far back as two or three years ago in ADSL at 1 1/2 megabits. And we are considering deploying the ADSL2 version which is the 6 megabyte version. And, of course, our SDV trial in Dover is still going along. So we are indeed proceeding with constructing broad band capable networks that will do high speed data as well as high speed video signals. MR. WERBACH: Let me ask the same question of David. A couple of years ago I think it was John Malone from TCI was on the cover of Wired saying I think end of '95 cable will be providing broad band to homes. That obviously hasn't really happened either, has it? MR. REED: I'm not going to speak for John Malone, but our -- it is happening. To be sure, there is a hype factor, perhaps an over hype. But to the extent something's new, that's a natural human reaction to things. You do have commercial high speed data services, but Time Warner in Akron, TCI in California, Cost in Phoenix and Comcast in Baltimore. And these are real. These are actual subscribers that are receiving this service. And we're learning from those commercial deployments. They're not technology trials. They are commercial deployments. MR. MAXWELL: Can we help to scale this? When we talk about 100 million access lines, it would be interesting to say by the end of '97, what would you guess to be the table modem penetration? By the end of '97, Pat, what would you expect to be the ISDN penetration in Bell Atlantic territory? MR. REED: I'd like to -- to be honest, we don't track that kind of information in terms of market forecast. It's just simply not something we do. And so I can't answer that. I mean, I can say that our member companies may have very aggressive roll outs. To Bob's question earlier about are there any regulatory implications to that roll out, we think that the two-way cable high speed data services is a nice service that roils out independent of the regulatory environment since it's just a network Internet access service. Some of our member companies are looking at what we call a telco return model which is because it takes time to activate all your cable network, you would have high speed access, high band width, in the downstream direction and the telephone line in the return. In that case, the extent to which anything that might impact how folks can access that telephone line in return could potentially be a barrier and that's a bridging strategy for the cable industry until they can do that. So that could slow down roll out if something like that were to occur. But I don't think referring to TCI, they hit the brakes on capital. I think it's a natural reevaluation process. It's not that they're not going to do any of these things. They're just reevaluating and figuring out how to do it more efficiently. And they'll be going forward and they are going forward. They have built a lot of state of the art 750 megahertz plan. And so you will see these services rolling out. But, for example, with high speed data service, it's associated -- it's not just the local access. You know, they have at home which working on trying to work on some of the other bottlenecks in the service. MR. PEPPER: Do you have any idea what percentage of the cable network today in terms of homes passed actually have been upgraded to the 750 megahertz so that they would at least be capable? Stagg's point we first have to do the hybrid fiber coaxial upgrade before you're then ready to go to the next step. Do you have an