SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : LAST MILE TECHNOLOGIES - Let's Discuss Them Here -- Ignore unavailable to you. Want to Upgrade?


To: Lance Bredvold who wrote (9920)12/24/2000 9:04:26 AM
From: Kenneth E. Phillipps  Respond to of 12823
 
Bypassing the online congestion

seattletimes.nwsource.com

by Alwyn Scott
Seattle Times business reporter

Start-ups like Microsoft come along only so often.

Keep an eye on one called InterNAP Network Services, a Seattle company out to change the way the Web works.

Remember the high-speed Internet? With streaming audio and video, it was supposed to remake the world, opening vast new uses for the humble PC.

But many people still find surfing at home a frustrating experience. Even at work, where connections are generally fastest, the wait for radio and TV feeds over the computer can be agonizingly slow, jerky and unreliable.

The problem isn't the superhighways. Phone companies are
spending millions to build fast networks that can zip Internet-data packets to their destinations.

The problem is the basic structure of the Internet. What started as a single network linking a few U.S. universities has grown into a network of networks, passing billions of data packets backand forth around the world.

Today, most people connect to just one or two of those
backbones though a service provider such as America Online or EarthLink. The roads are usually clear as long as the place they want to reach is on the same set of networks. If it isn't, the packets must travel through a series of narrow interchanges, known as peering points.

At rush hour, these onramps are as crowded and slow as
Interstate 5, with traffic backed up for miles.

This limitation has put the brakes on audio, video, interactive games and Internet-telephone services that rely on fast connections. Even many people who have ponied up the big bucks for Digital Subscriber Lines or cable connections find their performance less than thrilling because of slow interchanges.

InterNAP's founders saw the problem coming in the mid-1990s,
when they were installing Microsoft's first Internet connections and others around the region. To solve it, they began building a number of private onramps, called P-NAPs, that today connect to 11 of the biggest highway systems containing 95 percent of the world's Internet addresses. InterNAP's private ramps offer direct access to those highways, allowing users to bypass the congested
interchanges.

Tony Naughtin, InterNAP's chief executive, knew the Internet had a big future in 1994 when Microsoft, unsolicited, ordered a high-speed connection known as a T3 circuit. Such connections are common for businesses today, but back then were used only for national backbones. It was a defining moment.

"When Microsoft was stepping up and buying a large pipe to the Internet, they weren't doing it for a hobby," Naughtin says. "They were doing it for business."

Naughtin believed the Internet would become unavoidable and
began designing a better way for companies to connect. "We
came up with the idea that you could have a local hub that had connections to all of the major networks," he says.

Funding came from Paul Allen's Vulcan Ventures, among others, and InterNAP went public in September 1999. Its stock rushed to $52.97 the first day of trading, and peaked at $102.91 in February, but has fallen along with other Internet stocks. It closed at $6.31 Friday. Today, Amazon.com, Nasdaq and eBay are among those that use InterNAP to ensure that shoppers or traders on any of those 11 networks can reach their Web sites quickly and directly.

The system is too expensive for individual users, although a
number of Internet service providers, such as Speakeasy and
Seanet in Seattle, offer retail-price connections to P-NAPs. But for online businesses, the 10 percent premium InterNAP charges over commercial-connection rates is money well-spent.

InterNAP's software finds the best route between customer and client. And the choice of networks it offers ensures that a bottleneck on one highway won't cripple them - a feature online retailers take seriously during the crucial holiday season.

As with most Internet businesses these days, investors are
focused on how InterNAP makes money. Essentially, the
company buys and sells space on the big backbones. InterNAP
pays the networks for access to their highways. But because it is a big customer, it gets discounts of 30 percent to 50 percent, which, combined with the premium it charges customers, is its main revenue stream.

In addition, it offers businesses the option, for a fee, of putting their equipment in the same room with its P-NAP equipment, eliminating yet another link in the chain of connections to the Web.

InterNAP has set up 26 locations in major cities around the U.S. such as New York, Boston, San Francisco, Los Angeles and Atlanta. It also is building sites in London, Amsterdam and Tokyo, and plans to install about 12 more service points in the next year.

In Seattle, the company has opened a 16,900-square-foot space to house its P-NAP and customer equipment in high-tech Fisher Plaza, which, with its dual power sources and high earthquake resistance, offers the advantage of being virtually disaster-proof.

Other services, such as Internet security and filtering to weed out porn and hate sites, can be layered on top of the P-NAPs and offer potential additional revenue sources.

"We're like a reseller for all of these other networks," says Paul McBride, the chief financial officer, who thought up the idea of making the architecture pay.

Of course, InterNAP isn't alone in trying to improve the Internet's performance. Some phone companies are waking up to InterNAP's solution and discussing ways to monitor the service levels at interchanges points between their networks, says John Logo, vice president of data services at Global Crossing, which is building one of the major Internet backbones. Already, some of them see InterNAP not just as a customer, but as a competitor as well.

"To the extent that customers buy connectivity from them as
opposed to Qwest, they are a competitor," says John Morgan,
director of Internet-solution strategy at Qwest Communications, the Denver-based phone company whose territory covers much of the West, including Washington state. "To the extent that we're all going after end-user customers, they are a competitor."

But InterNAP contends that network operators don't have much
incentive to improve. At present, data traffic doesn't actually generate much money for the phone companies. Most revenue comes from carrying voice-telephone calls, which use different technology. As networks are converted in coming years to the cheaper system of carrying voice calls over the Internet, however, competition to improve performance will increase.

And other start-ups also are trying to build a better Internet mousetrap, too. Companies like Akamai Technologies use servers scattered around the Web that hold caches of frequently accessed pages to make them pop up more quickly. But because those services are stored, they can't improve voice or other real-time communications.

Some contend that by bundling the major networks into a single package for customers, InterNAP's solution - which the company has patented - changes the basic architecture of the Internet.

"At this point in time they're the only company out there doing this kind of business," says Kevin Giboney, a broadband-communications analyst at D.A. Davidson, a regional investment company based in Great Falls, Mont. "InterNAP has been pretty much flying under the radar. They are a young start-up, which in terms of revenue hasn't had much impact."

But the solution, he says, points up weakness of the Internet's current configuration. "And a lot of the newer service providers that are coming on - Intel, Digital Island, 360networks, Enron, Akamai, Allied Riser, Inktomi, Adero - are using InterNAP as a solution," Giboney says. One example: Under an agreement this month, Akamai will pay a fee to place more of its servers in InterNAP facilities, while InterNAP will resell Akamai service through its hubs.

InterNAP's results are improving. It said this month fourth-quarter results will surpass analysts' expectations and that revenue in 2001 will reach $220 million, compared with earlier estimates of about $202 million, and that it will break even two quarters earlier than forecast.

Phone companies still insist their model is stronger. Owning the networks, Qwest's Morgan says, allows them to offer lower prices and better service than any reseller. But InterNAP avoids the huge capital cost of building and maintaining the networks. And it notes that because of fierce competition, phone companies don't really want to help each other improve data traffic.

Indeed, the problem with those interchange points is that when the Internet was being set up, phone companies agreed to pass the data traffic to each other free. This meant a phone company had an incentive to push off its network as quickly as possible any data not intended for one of its customers. The result was a big game of hot potato, in which traffic is simply dumped at the nearest interchange with the nearest network, regardless of whether that's where it's trying to go.

Some of the phone companies say they are taking steps to
improve the relationship to maintain certain levels of service. "We work with our peers to make sure we have unobstructed interconnection between all our peers," says Global Crossing's Logo. "We believe that model is going to shift. We've had discussions with some of our peers in moving toward service-level agreements across the peering relationship."

But antitrust concerns prevent the major phone networks from
cooperating too closely. In a recent paper, the Federal
Communications Commission said InterNAP's solution provides a welcome third hand in the marketplace that allows Internet traffic to travel between many networks.

"Clearly there's a value for what they do now," says Brent
Bracelin, an analyst at Pacific Crest Securities in Portland. "The question is, is it a niche service, or do they become this core enabler of the Internet?"

Copyright © 2000 The Seattle Times Company



To: Lance Bredvold who wrote (9920)12/24/2000 3:43:08 PM
From: axial  Read Replies (1) | Respond to of 12823
 
Hi, Lance -

I'm not sure of the particular conditions in Korea, but in the current US environment, IS-95 provides an operator 12.5 MHz of spectrum, which will give 57 channels, assuming a three-sector cell site.

The beauty of CDMA is that the 58th caller can still be accomodated, (or handed off from one site to another), by using the so-called "soft degradation" feature of CDMA; ie., an increased bit-error rate will allow the extra users.

IS-95B provides the following features and improvements:
1 - efficient handoffs (CDMA-AMPS,CDMA-GSM,CDMA-CDMA)
2 - longer battery standby times (~two hundred hours)
3 - medium-rate data services (65-144 Kbps)
4 - improved capacity, (roughly 20%) coverage and efficiency

However, data transmission in IS-95B has a cost:

'The IS-95-B specifications have been written so that manufacturers of subscriber units and infrastructure equipment can potentially support data rates of up to 115 kbps. This leap in data speeds is attained by aggregating up to eight CDMA traffic channels for data transmission (8 x 14.4 kbps = 115.2 kbps). While data rates of 115 kbps will be possible with the IS-95-B standard, Qualcomm anticipates that operator requirements will initially warrant support of data rates between 28.8 kbps and 57.6 kbps on the forward link, and 14.4 kbps on the reverse link.

americasnetwork.com

So, it appears that it's the usual good news/bad news game.

Yes, you're going to have 20% increased capacity with IS-95B, but aggregation of channels for data transmission will eat into system capacity.

That's only my opinion, based on what I've been able to learn. I welcome correction by many others who know more.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

On the larger question of the promises of 3G there's no doubt that the proposed figures are deceptive. All of the high-end figures are a best-case scenario, that degrades as more users come online, and the network attempts to accomodate both data and voice.

What's going on here? Given population and usage growth in major urban centers worldwide, how long will it be until we're right back where we started, with full pipes (spectrum). Picture this: 10 years from now, when there's 4G in NYC, and everyone's trying to phone home because there's a snowstorm, will you be able to download a multimedia clip, at rush hour?

Not likely. You'll be lucky if you can phone home, just like now.

I tend to agree with Ray that the marketing efforts have obscured the reality. The only caveat I might add is that in off-peak hours, the systems will be able to supply some pretty good speed/data transmission.

The only answer that I can see, aside from VOIP, new technology, cellular architecture improvements, and other tweaks, would be more spectrum.

JMO ;-)

Jim