Content on the Internet is evolving rapidly, from simple, static web pages to complex graphics and audio/video streams. Delivery of this new class of content requires significantly more bandwidth. In the past the Internet was mainly used for e-mail and basic information retrieval, i.e. point-to-point exchanges between a user and a host server. But new applications are emerging.

For example, online learning and live multimedia webcasts require the ability to reach out to millions of users simultaneously with rich, bandwidth-intensive content. These applications require point-to-multipoint exchanges, also known as multicasting.

This emerging demand for content will further stress Internet infrastructure in terms of bandwidth, scalability and predictable content delivery. The current Internet infrastructure is not designed to efficiently deliver this new multimedia-rich content or to support the multicasting needs of emerging applications.

The traditional Internet is a point-to-point network, built to handle the movement of data from one point to another using the Internet Protocol (IP) standard. When the same data is sent to several locations, this is done by sending additional copies of that data, separately, from the sender to the individual receivers.

As the number of simultaneous receiver locations runs to the millions, as may soon happen with live events, webcasting and other content-rich applications, the existing Internet infrastructure will be unable to meet the bandwidth demands.

The traditional Internet is also what is known as a "best-effort" network. The path between each user's personal computer and the computer that holds the requested content is a series of interconnected computer networks. The content itself is comprised of many data packets that move from one network to another in a hop-by-hop manner.

Not all networks have the same capacity to carry data, and not all devices that interconnect those networks can process the data at similar speeds. If there is congestion at any of the intermediate hops, as is the case when a large number of users try to access the same content, some data packets can be arbitrarily dropped or delayed.

However, these solutions do not address the fundamental need for emerging applications to deliver media-rich, data intensive content to a large number of users simultaneously, in a scalable and predictable manner.

For example, recent attempts to deliver streaming audio and video content over the Internet to many listeners and viewers simultaneously, such as the Victoria's Secret webcast, the NetAid concert, the Star Wars: The Phantom Menace movie trailer release and the Super Bowl 2000 webcast, ran into significant difficulties in delivering a uniform and high-quality experience. In the aftermath of each of these attempts, it was widely reported in the press that many Internet users could not gain access to the content and that those who did received poor quality, unsynchronized audio and video.

The current Internet infrastructure, designed to facilitate interactivity and communications between users in a network, is often referred to as the Transaction Internet. The Transaction Internet most often experiences congestion when delivering event-based or streaming content over the existing infrastructure.

In addition, over the last five years many of these networks have migrated from analog to digital transmission systems, thus greatly enhancing their ability to carry new types of content. The ability to offer Internet-enhanced broadcast content would create new revenue opportunities for the service providers that own and operate these networks.

However, broadcast service providers face enormous technological challenges in extending their networks' capability to deliver Internet content as well as traditional broadcast programming. One such challenge is finding the bandwidth to offer Internet services in addition to traditional programming, because the latter typically consumes substantially all of their available bandwidth.

Current digital broadcast technology does not let broadcasters make full use of their transmission bandwidth. As a result, somewhere between two and ten percent of bandwidth goes unused, filled instead with empty placeholders called null packets.

Moreover, current digital broadcast network infrastructure is not designed to efficiently handle the delivery of Internet content. Digital broadcast service providers typically operate networks that are optimized for the one-way transmission of television-oriented services using the Moving Picture Experts Group (MPEG) standard; they lack a return path for data communications from users. And they have only very limited ability to integrate data seamlessly into their current audio/video streams.

Legacy equipment and services predominate in broadcast networks. Current networks were built using systems that are predominantly closed and proprietary. They limit the ability of broadcast service providers to use best-of-breed equipment in developing and promoting new services. Furthermore, broadcast service providers have an installed base of equipment that limits the changes they can make to their networks and requires any incremental changes to be completely compatible with existing equipment and services.

A new category of networking products is therefore required to overcome the challenges facing both the Internet and broadcast networks. These products will enable broadcast networks to connect with the Internet, creating a "Broadcast Internet" that seamlessly combines the powerful benefits of both.

This is where SkyStream Networks (SSNW:NASDAQ) comes in. It combines the efficiency, scalability and performance of broadcast networks with the interactivity and personalization of the Internet. Its products and technology enable broadcast networks to connect with the Internet, offering a cost-effective, reliable and seamless path for delivering rich content to traditional Internet service providers and end users.

By enabling the Broadcast Internet, SkyStream allows content to be delivered across multiple paths. Its products include networking systems known as media routers and sophisticated software that manages Internet content and services across the Broadcast Internet.

The Yankee Group, a global market research firm, has forecasted that the market opportunity for equipment that can enable this combined network will grow from virtually nothing in 1998 to approximately $3.7 billion by 2003. As you can see, SkyStream has a huge opportunity to be a big player in a new and evolving market.

Cable, satellite and digital television broadcasters can use SkyStream's products to offer multicast high-speed content or interactive Internet applications. With its products that use the MPEG standard, broadcasters can combine digital video streams based on MPEG and data based on IP in a cost-effective and seamless manner.

Internet service providers, telecommunications companies and content distributors can utilize SkyStream's solution as an incremental high-speed backbone for delivering content and services to the end user.

SkyStream also provides edge media routers that enable Internet service providers and telecommunications companies to remotely receive Internet protocol data services from broadcasters and deliver them throughout their networks. Its family of Broadcast Internet management software enables broadcasters to aggregate, schedule and broadcast web and data services, or channels, to a target set of receiving locations on their network.

Direct sales efforts have been primarily focused on satellite, digital television and cable broadcasters. For the year ended December 31, 1999, direct sales accounted for approximately 82% of revenue.

Indirect sales include an original equipment manufacturer agreement with Harris Corporation, a leading supplier of broadcasting equipment to the television market. Harris Corporation is the largest manufacturer of digital television equipment in the United States.

Revenues from its five largest and ten largest customers accounted for 66% and 77%, respectively, of its total revenues for the year ended December 31, 1999. EchoStar Communications accounted for 38% of revenues in 1999, International Datacasting Corp. for 12% and Cablecom for 10%. This is risky business, as the company has no long-term contracts with any customers for the purchase of its products, and customers may reduce or discontinue purchases at any time. Customers may also delay or cancel purchase orders without penalty.

SkyStream began shipping its first product, the source media router, in February 1998. Prior to this, operating activities were mostly devoted to research and development.

Revenues for 1997 were $71,000. Revenues increased from $1.3 million in 1998, SkyStream's first year of volume shipments, to $8.5 million in 1999. Four customers accounted for approximately 75% of the revenues in the year ended December 31, 1998, and four customers accounted for approximately 64% of revenues in the year ended December 31, 1999. After the launch in 1998, revenues have been steadily increasing — a good sign.

International sales accounted for 38% of its revenues for the year ended December 31, 1999, and are expected to increase in the future.

Companies such as Cisco Systems, Lucent Technologies and Nortel Networks have historically dominated the Internet infrastructure market, while the broadcast network equipment market has been led by firms like General Instrument, Harmonic and Scientific-Atlanta.

Let's look at the performance of some of SkyStream's competitors to get an idea of the potential the company has in this industry. Cisco Systems (CSCO:NASDAQ) is no stranger, with a 52 week high of US$80.06. Lucent Technologies (LU:NYSE) had a 52-week high of US$82.28.

The CFO has been with the company since October 1997 — first serving as director of finance, then as vice president of finance and now as CFO. Prior to joining SkyStream, she served as director of finance for Target Therapeutics, Inc., a medical device company acquired by Boston Scientific in 1997.

The vice president of engineering and chief technology officer joined the company in March 1998, first serving as vice president of manufacturing and engineering. Before coming to SkyStream he held various positions with 3Com Corporation.

Directors include a couple of venture capitalists — something to keep an eye on. Venture capitalists tend to spook me a bit because they sometimes care more about their own interests than about the well-being of the company — a simple matter of survival of the fittest.

One of the company's directors is also a general partner of the Mayfield Fund, a venture capital firm. He is also a director of Advent Software, Inc., a portfolio management software company, Montgomery Street Income Securities Inc., a closed-end bond fund, and Netcentives Inc., an Internet marketing, products and services company.

Another director is a general partner of Institutional Venture Partners and a managing director of Redpoint Ventures, both venture capital firms. He is a director of Turnstone Systems, Inc., a provider of digital subscriber line deployment and management products, MMC Networks, Inc., a developer of network processors, Ask Jeeves, Inc., a natural-language question answering service on the Internet, and TiVo, a personalized television services company.

SkyStream Networks, Inc. has a strong lead underwriter, Goldman, Sachs & Co. The proposed ticker symbol is SSNW and it will be listed on the NASDAQ.

For more information, please contact SkyStream Networks, Inc. at 555 Clyde Avenue, Mountain View, CA 94043, phone 650-390-8800, fax 650-390-8990 and website.