|
|
 EE Times is the leading
weekly newspaper for design engineers and technical managers who are major
decision-makers in the electronics industry. The publication is circulated
to professionals working in the computer, communications, industrial,
military/aerospace, components, test & measurement, and other industrial
OEM markets.Visit EE Times Online for more technology news and features. |
|
The Network Game The network game is one of big risks and big rewards. Universal PC-slot plot Upgradable-PC effort takes divergent paths. Wireless Dedication Raytheon steps into wireless LAN market. Microtec's mega move Post-merger Microtec widens embedded focus. Amati in Japan Sumitomo to sell Amati ADSL gear in Japan markets. Rambus IPO tips Intel links Yakatori. The Rush to Russia Would-be investors in Russia advised to be wary. |
|
By David Gardner, EE Times Boston -- At the end of 1993, Cisco Systems Inc. invested $2.7 million in startup Cascade Communications Corp. For the networking colossus, the idea was to have the inside track to acquire Cascade if the new company became a success. As it turned out, Cascade was very successful indeed. But rather than making Cascade acquisition bait for Cisco, the $2.7 million helped ignite a chain reaction of furious mergers among networking companies. Eventually, Cascade jilted Cisco. Then, two weeks ago, it jumped into the waiting arms of Ascend Communications Inc., whose mounting challenge to Cisco is more potent today thanks in large part to Cascade's robust product portfolio. The value of the deal, whose initial bid was $3.7 billion in Ascend stock, fluctuates as the stock's price rises and falls. In other words, the networking acquisition game is fraught with big reward as well as big danger. The acquisition of startup networking companies with hot products--often ATM products--has enabled the galloping networking market to develop at breakneck speed as it attempts to keep up with the Internet explosion. The feeding frenzy aspect of networking acquisitions is a relatively new phenomenon and to a large extent it has replaced traditional and time-consuming methods of bringing products to market. "The issue is time-to-market," said Bob Machlin, vice president of marketing at Cascade, explaining that it used to take at least three to five years to develop new mainframe computer products. Today, sophisticated network infrastructure products are often developed in less than one year. Noting that Cascade itself has paid $150 million to $250 million for startup companies to acquire products, Machlin added: "It [acquisitions] is almost not an issue if you want to be in the market." Thus, the phenomenon of a uniquely exploding market demand for leading-edge products situated largely in the slow-moving and conservative carrier business created a very combustible networking acquisitions market. Said Machlin of Cascade's users: "They want bigger, stronger vendors to buy from." A key reason the acquisitions have been important is that larger companies have relied on them to flesh out their product lines rather than spend on the traditional but lumbering research and development approach, thereby relying on acquisitions to keep their forward momentum. Cascade, for instance, made several acquisitions before being acquired by Ascend. In turn, Ascend will likely be an acquisition candidate some day. But now, as the stocks of the major networking companies have plunged in recent weeks and as prominent insiders have safely sold out stock positions at high prices, the burning question is whether the networking merger binge is over or whether the acquisition wave is just catching its breath before beginning a new onslaught. "As the networking stocks go lower and lower, I think it will bring in the larger companies," said Brendan Hannigan, senior analyst at Forrester Research in Cambridge, Mass. "The stock multiples are getting back to reality. But I don't think the acquisitions will slow." Hannigan explains that larger companies like IBM and AT&T have low stock price multiples, which make it difficult for them to use their stock to acquire smaller firms. Networking companies have been using their high-flying stocks to purchase smaller companies. At the top of the networking pyramid, Cisco, the largest, has played the networking game successfully, most notably with its $4 billion acquisition last year of Stratacom. With a $35 billion market valuation, it hasn't been a struggle for Cisco to set aside $1 billion in an acquisition fund to buy a few companies every year. But even Cisco can stumble. It couldn't make a deal with Cascade in 1995, when it was said to have offered $1.6 billion for the smaller company, and its acquisition of Lightstream never resulted in much in the way of products. Bay Networks Inc.--the combination of Synoptics Communications Inc. and Wellfleet Communications Inc.--is an example of a networking merger gone wrong, although a new management team is currently working to get Bay Networks back on track. (Larger than Cisco in 1993, Bay Networks today is less than one-third Cisco's size as the Wellfleet-SynOptics combine turned out to be synergism in reverse.) The biggest networking merger of all--3Com Corp.'s $7 billion planned acquisition of U.S. Robotics Corp.--has been greeted by a collapse of both companies' stock. Cabletron Systems Inc. is another major player in the networking acquisition competition. As befits the networking moniker, the companies in the field appear to do a great deal of it themselves. Just as Ascend and Cascade did the networking mating dance, so do hundreds of other outfits. Everybody seems to be talking to everybody else in the business, and investment bankers and assorted deal makers are prowling the industry constantly on the lookout to broker an acquisition. Ascend and Cascade were said to have been talking for two years. Last summer, Cascade founder Gururaj "Desh" Despande talked openly of merger possibilities, and industry savants immediately linked the talk with Ascend--with or without any evidence. As recently as early January Ascend's president, Mory Ejabat, noted publicly that the two companies talked frequently, but weren't talking merger at the time. Then the companies got serious; in late March, Ascend announced the offer to buy out Cascade. So far as acquisitions go, Cascade is particularly interesting because it is both the acquired and the acquirer--as Cascade is preparing to be digested by Ascend, Cascade is still digesting Sahara Networks Inc., which it picked up in January. Last May, Cascade acquired Arris Networks Inc., a nine-month old company with a hot ATM switch under development. The Arris startup is a prime example of the new breed of ATM developer: it is just a few months old, has a hot ATM product in the pipeline and has no profits and no sales. Typically, the hot product is still in the developmental stage. It is hoped that it will one day function as advertised. In its acquisition of Arris, Cascade didn't have to look far or check out the curriculum vitae of Arris' management--Arris was located nearby in Westford, Mass., and its founder, Wu Fu Chen, had been a founder of Cascade where he had served as chief technology officer and vice president until he left it to form Arris. The financials were good, too: both Cascade and Arris had been funded by networking venture capital powerhouse Matrix Partners of suburban Boston. Last May, Cascade acquired Arris, its 25 employees and its ATM developmental product for $145 million in stock, which was a nice hit for nine months' work. By all accounts they have been a rousing success. The AX 800 and AX 1600 High Capacity Access Switches are both in full production and several are in the hands of customers. The AX family was designed with service providers in mind and the family combines ATM and Frame Relay switching with switching capability for integrating analog modem-based applications with high-speed digital services like ISDN and leased lines. The Model 1600 supports up to 672 integrated digital modems in a single shelf. A few months after funding the Chen team, Matrix Partners bankrolled another ATM team from Cascade headed by former Cascade lead engineer Bing Yang. The scenario was repeated in similar fashion: 27 employees, a hot ATM product in the pipeline and an acquisition of the startup, Cadia Networks, for $150 million. This time the acquirer was Fore Systems Inc. (Warrendale, Pa.). Fore has been in the vanguard of the ATM LAN market and Cadia Networks is developing a ATM WAN product aimed at ISPs that is scheduled to be shipped in the second half of 1997. Once again, the multimillionaires led by the 34-year-old Yang were tripping all over themselves. But Cascade's most exciting acquisition is Sahara, a General Datacomm spinoff that has been titillating networking users ever since the announcement in February 1996 of its SaharaView architecture for building and managing ATM nets. Last summer Sahara unveiled a three-product family of ATM muxes aimed at the WAN market. The products are managed by SaharaView, which, like many of Cascade's products, is said to be easy to customize. Another benefit of the family is that it can be viewed through any Java-enabled Web browser. Originally scheduled for delivery late last year, Cascade's Machlin said the Sahara products should be in beta test during this quarter with general availability set for the third quarter. But can companies like the six-year-old Cascade, with 1996 revenues of $341 million and 1,000 employees, start up in today's networking environment? "Not in our space," said Machlin, who believes the new networking startups will be in "uniquely focused niches." Forrester Research's Hannigan thinks attractive ATM entrepreneurial opportunities in networking will be in WANs rather than LANs and he cites a recent Forrester survey of networking users. "ATM in the LAN is just not going to happen," said Hannigan. "We think fast Ethernet and Gigabit Ethernet will take over the LANs. It's pure economics." The only Massachusetts high-tech company that grew faster than Cascade in so short a time--Lotus Development Corp.--was also acquired when IBM picked it up less than two years ago for $3.52 billion. The Lotus acquisition was followed by a mass exodus of top managers, but IBM has been crowing that the acquisition has been a great success with the Lotus Notes groupware user base more than doubling last year alone. In the networking acquisition world, IBM's acquisition stands as a prime example of a networking merger gone right. Ascend and Cascade are also moving to ensure that their merger works right, too. And the fact that both companies are entrepreneurial in nature should help. Ascend's Ejabat will be the CEO and president of the surviving company, headquartered in Alameda, Calif., while Cascade's CEO Daniel Smith will head up the switching-products unit in Westford. The new combines will enable Ascend to bring market leader Cisco into its gunsights as networking competition grows. As for Cisco, its failure to acquire Cascade doesn't appear to bother it one bit. Its $2.7 million investment in Cascade just over three years ago was being sold off at more than $276 million recently--a gain of more than 100 times on its original investment. (c) CMP Media, Inc [This article comes from EE Times in a joint cooperative effort with the Motley Fool. For more articles like it, please look at Fool's Gold every weekend or simply go to the Fool's Gold Mine and page through our back issues, which all have clever and cool EE Times articles in them.] |
|
Upgradable-PC effort takes divergent paths By Rick Boyd Merritt, EE Times Houston -- Creating a more modular, upgradable and consumer-electronics-like personal computer seems to many designers to be a worthwhile goal. But just how to do that is a matter of some debate. An ad hoc group driven by Compaq Computer Corp., Intel Corp. and Microsoft Corp. recently released preliminary details of planned specifications for what it calls Device Bay, a universal slot into which a wide variety of computer add-ons could be plugged. But tiny startup NeoSystems Inc. has gone public with a broader concept of modularity that it hopes to license to computer makers and, ultimately, to consumer-electronics companies. The Device Bay group claims that as many as 60 companies, including its heavyweight organizers, will back its specifications for desktop and mobile add-ons, which are expected to surface in the second half of next year. "Some OEMs see Device Bay as a challenge to their ability to differentiate their systems, but I think they also see the potential pot of gold at the end," said NeoSystems sales director Brian Welsh, who recently took part in a closed-door meeting on Device Bay that was attended by about 170 people. "The thing they need to see is the path from the business they have today to the advantages Device Bay could bring long term." The Device Bay group has devised what it claims is a simple, unique connector that can use the Universal Serial Bus (USB) or the IEEE 1394 interface to let users plug add-ons into PCs as easily as they load tapes into VCRs. Device Bay will accommodate such add-ons as hard disk drives, CD-ROM drives, DVD players, high-density floppy drives, smart-card readers and such new device types as videoconferencing adapters and digital-broadcast-satellite tuners. Meanwhile, NeoSystems, has developed an approach for upgrading nearly every piece of a system, including the main CPU. The startup's approach uses a standard Peripheral Component Interconnect (PCI) card physical interface, over which signals can be sent for such interfaces as Enhanced IDE, PCI, ISA, USB and IEEE 1394. One point on which the two camps clearly agree is that the rapidly evolving PC needs a more user-friendly way to expand. "Today, the upgrade market is held back by the hassle, expense and fear of getting inside that box," said Joe Pajer, director of product planning for Compaq's Commercial Systems Division. "By eliminating those issues, we will revitalize the upgrade market." Device Bay "will cut across commercial and consumer desktops, NetPCs and mobile systems," Pajer added. "Almost every segment will want at least one Device Bay, and some may want two or three in a system." The Personal Computer Memory Card International Association (PCMCIA) intends to take a close look at what is today a draft Device Bay spec for mobile systems. "We've decided to go ahead and establish a beginning standards effort, and we will invite the Device Bay people to take part in it," said Charlie Payne, president of PCMCIA. Payne expects the PCMCIA to adopt a Device Bay interface but said he suspects that "some accommodations [to the Device Bay mobile systems. My impression is very little has been done on the mobile Device Bay specification to date." A mobile universal-bay group will formally be established at the next PCMCIA meeting, slated for mid-May in Paris. If that group decides to modify the Device Bay work for mobile applications, the changes could be completed and the spec ratified about nine months later, Payne estimated. For its part, the Device Bay group has already issued at least two proposals for mobile-systems support, and it hopes to make its entire specification public before June. (c) CMP Media, Inc [This article comes from EE Times in a joint cooperative effort with the Motley Fool. For more articles like it, please look at Fool's Gold every weekend or simply go to the Fool's Gold Mine and page through our back issues, which all have clever and cool EE Times articles in them.] |
|
Raytheon steps into wireless LAN market By Loring Wirbel, EE Times Andover, Mass. -- After several years as a component-level power behind the throne in wireless-LAN markets, Raytheon Wireless Solutions is launching its own 2.4-MHz frequency-hopping product for horizontal computing markets. The creation of a dedicated wireless-LAN infrastructure group within Raytheon Microelectronics signals a move toward commercial end users for a company that has been returning to its government roots. Its parent company, Raytheon Co. (Lexington, Mass.), has made on-and-off efforts to enter internetworking OEM markets through acquisitions, but in recent months the company has turned back to its original home in defense and government electronics. Several LAN-related businesses were sold off, while the acquisition of defense divisions from Hughes Communications Inc. and Texas Instruments Inc. will bring the parent to more than $20 billion in annual revenues when completed. But the Raytheon Microelectronics group is not a stranger to commercial uses of wireless LANs. It worked closely with Windata Corp. and other early pioneers in integrating DSP modems, medium access control (MAC) devices, and RF/IF components in proprietary solutions for the FCC's unlicensed ISM (Industrial, Scientific & Medical) bands. The heart of Raytheon's expertise is the gallium-arsenide RF and microwave monolithic circuits that have played a key role in the Iridium space-based PCS system, and other wireless systems. Roland Fournier, product manager for the new Raylink WLAN family, said that DSPs, MAC-layer logic and RF/IF components all were developed internally because "our frequency-hopping solution required tight integration among baseband, MAC and analog portions of the design." For the time being, these components will only be offered as part of PC card and access-point system-level products, though Raytheon may consider chip-set sales or licensing at a later date. Although Raylink might be entering the market late by some analysts' reckoning, the slow pace of the IEEE's 802.11 standards group (a final standard is expected this summer, following more than five years of work), and the low uptake of wireless LAN technologies in many horizontal office applications, has relegated existing players like Proxim Inc. and Lucent Technologies' WaveLAN group to vertical markets in retail point-of-sale and factory-floor applications. Fournier said that Raytheon is betting the time is right to introduce wireless LAN products to the horizontal market, to be sold as a complement to wired LANs rather than as a replacement. Raytheon will look for some vertical applications in health-care fields, but the primary target market is mainstream office applications. (c) CMP Media, Inc [This article comes from EE Times in a joint cooperative effort with the Motley Fool. For more articles like it, please look at Fool's Gold every weekend or simply go to the Fool's Gold Mine and page through our back issues, which all have clever and cool EE Times articles in them.] |
|
Post-merger Microtec widens embedded focus By Terry Costlow, EE Times Santa Clara, Calif. -- Microtec Research Inc. has been a sleeping giant in the embedded-software and -tools environment since being acquired by Mentor Graphics Corp. a little more than a year ago, but its strategic plans are now complete, and the company is unveiling products that it believes will move it solidly forward. The acquisition by Mentor, which was closed at the start of 1996, gave Microtec the deep pockets of a large corporate parent. That's sharply different from the status of Microtec's competitors, most of which are comparatively small public and private companies. The acquisition by Mentor drew industry interest because Mentor was expected to bring a stronger management focus to a company that had been struggling with its corporate identity. About a year before it was acquired by Mentor, Microtec made a move to step beyond its tools-supplier base when it acquired Ready Systems. That acquisition moved Microtec into the real-time operating system (RTOS) market, but buying a company of nearly the same size also blurred Microtec's focus. "Microtec was a small company that had gone through a merger [with Ready] a year before, so when [Mentor Graphics] showed up there was no clear strategy," said Microtec's chief operating officer, Vin Ratford, who came to the company from Mentor. "Their stated business plan was to better serve the telecom industry, but it was difficult to get a tangible grasp of what that meant. The merger gave us time to go off and decide what the strategy was and to figure out how to best implement that strategy. We developed a well-thought-out three-year strategic plan and a tactical plan for the first year." The joint strategy-development team decided that the focus on telecom should be broadened to related markets such as data communications. The new strategy and implementation plan also recommended an examination of the ways that the companies' various products worked together, and to leverage those strengths. Microtec had previously had a point-product strategy, Ratford said. In fact, after the Ready-Microtec merger, this product-oriented strategy was quite sensible, he said. Now Microtec has established both product differentiation and a stronger corporate identity. Though its tools work with the VRTX real-time operating system, they aren't tied to it alone. "What makes us different is that we're the only company delivering tools that support more than one operating system," Ratford said. "Telecom companies typically deploy multiple operating systems with proprietary software, and often several off-the-shelf operating systems." (c) CMP Media, Inc [This article comes from EE Times in a joint cooperative effort with the Motley Fool. For more articles like it, please look at Fool's Gold every weekend or simply go to the Fool's Gold Mine and page through our back issues, which all have clever and cool EE Times articles in them.] |
|
Sumitomo to sell Amati ADSL gear in Japan markets By Ron Wilson, EE Times Denver -- Sumitomo Corp. of America has signed an agreement that aims to make digital-subscriber-line (DSL) modem equipment available to the Japanese market. Sumitomo will distribute throughout Japan the asymmetric DSL subscriber modems and central office equipment produced by Amati Communications Inc. (San Jose, Calif.). In the first phase of the relationship, Sumitomo and Amati will jointly distribute Amati's ADSL equipment based on discrete multitone coding. In the future, Sumitomo will distribute Amati hardware based on the emerging very high-bit-rate DSL (VDSL) standard. (c) CMP Media, Inc [This article comes from EE Times in a joint cooperative effort with the Motley Fool. For more articles like it, please look at Fool's Gold every weekend or simply go to the Fool's Gold Mine and page through our back issues, which all have clever and cool EE Times articles in them.] |
|
Yakatori By David Lammers, EE Times Tokyo -- Rambus Inc.'s planned initial public stock offering has shed light on the sensitive issue of how it will work with Intel Corp. while also striving to broaden its customer base. The IPO is expected before the end of July. Three million shares will be added to more than 17 million now held by employees and venture-capital backers. Don't rush out to buy unless you are a major client of one of the bankers or brokers, or are otherwise part of the privileged minority. Intel, by dint of its agreement to back the Rambus architecture, has the right to buy a million shares at $10 each, though there are some interesting quid pro quos. For one, Intel can exercise the warrants only after at least 20 percent of Intel's chip sets are shipped with the Rambus interface logic. Although that sounds like a clever way of making sure Intel remains committed to the Rambus architecture, the financial ties mean relatively little to Intel. The agreement also keeps Intel from using Rambus patents to develop a logic interface to other high-bandwidth DRAMs that might become tempting, such as the Synclink DRAM or double-data-rate (DDR) synchronous DRAMs now in development. More importantly, the agreement forbids Intel to get between Rambus and microprocessor competitors such as AMD and Cyrix, or chip-set vendors, should they choose to develop a Rambus-capable interface. Now that Intel is committed to using Rambus DRAMs and the MPU-to-memory Rambus channel to market by 1999, Rambus obviously wants to get the rest of the computer industry on board as well. Intel also made clear it won't tolerate Rambus shenanigans. Though Rambus doesn't disclose initial license fees or per-chip royalties, it has been clear that early Rambus backers like NEC and Toshiba got a better agreement than the DRAM companies that will sign up from now on. The agreement bars Rambus from charging more than a 2 percent royalty after the RDRAMs start being shipped in volume as PC main memory. Rambus could ask for a 2.5 percent royalty from a latecomer, but the royalty must drop to 2 percent or less once the PC industry adopts RDRAMs. Finally, though Intel may contribute technology to the 1.6-Gbit/second Direct Rambus version, Rambus controls that specification and has the rights to license it. Moving to a 16-bit-wide interface, to a 400-MHz clock speed with double data rate, and improving the protocol, will deliver video data rates to the desktop. That is what will keep Intel happy, no matter how valuable its Rambus shares become. The only question remaining is how fast DRAM makers can master the hidden intricacies of the Rambus interface, and get their yields up and prices down to the levels of today's memory solution. (c) CMP Media, Inc
[This article comes from EE Times in a joint cooperative effort
with the Motley Fool. For more articles like it, please look at Fool's Gold
every weekend or simply go to the Fool's Gold Mine and page through our back
issues, which all have clever and cool EE Times articles in
them.] Would-be investors in Russia advised to be wary (Second of two parts) New York -- The appeal is undeniable: some market analysts see the Russian telecom industry as the standout portfolio investment of 1997. In fact, GIST Telecom Index, which tracks share values of 15 Russian public-network operators, "was up more than 76 percent in mid-March 1997 compared with the beginning of the year," reported the March issue of a newsletter published by Global Information Services and Technologies Inc. (Arlington, Va.). But before Western investors check their atlases for the location of Yakutsk and Smolensk, Americans and Western Europeans who have spent time in Russia say they should be aware of what to expect. Doing business in Russia is very different from doing business in the United States. "It takes a lot of time, effort and persistence to learn all you need to know financially and legally," said Christopher Worth, a partner with the accounting firm Coopers & Lybrand. "But it's worth the effort." Dazzled by ample opportunities to make money, companies eager to invest in Russia's booming telecom industry must also take the drawbacks into consideration. These include sometimes-chaotic financial and legal practices, bribery as a cost of doing business and infrastructure issues. As in many emerging markets, telecom expects say, the industry's infrastructure is in poor shape, requiring massive levels of investment to upgrade existing lines and install new ones. While the state and other domestic sources will provide some of the necessary funds, it is expected that much of the money will have to come from large international telecommunications companies seeking access to the Russian market. Beyond infrastructure, old Russia hands say newcomers are sometimes caught off guard by the difficulty of obtaining information that Western companies take for granted. For example, investors want to quantify a company's assets and financial performance, but executives who have spent time in Russia say it is hard to get an adequate financial picture of Russian companies. "A key thing to remember is that financial measures have traditionally played little or no role in the management of these companies, and key performance indicators are not available," said Worth, who has spent time working there. In addition to having difficulty getting accurate financial data on Russian companies, U.S. companies investing there must be aware of risks associated with the Russian legal system. "There are inconsistencies among laws, conflicting local rules and regulations, lack of administrative guidance on legal issues," warned Jeffrey Burt, a partner at Arnold & Porter. "And judges and courts are relatively inexperienced at interpreting laws." In such an environment, Burt said, U.S. firms and investors can experience frustration and disappointment--and even financial losses--unless they know the ropes. Indeed, some telecom companies have been burned over seemingly mundane things like acquiring the government licenses they need to provide telecommunications services. These licenses are issued and renewed for periods ranging from three to 10 years, said Burt. Sometimes a license is issued quickly; sometimes it takes months. Moreover, several different licenses may be issued to one person, but these may not be transferred. Failure to abide by this law could result in the license being revoked or suspended. A separate fee is required for each license; so is a mandatory contribution to finance the development of the public switched telecommunications network in Russia. Another area where some foreign companies get into trouble is with telecom equipment. A Russian communications law requires that all gear must comply with standards and technical requirements established by the Russian government. In addition to laws specific to telecommunications, said Burt, it's important to remember basics about Russia's legal system. Most important: U.S. court decisions are not enforceable in Russia, but U.S. arbitration decisions are. Burt said American companies must also accept that some business practices in Russia are unsavory. For instance, bribery is common, and a company may find it expedient to bribe a government official to get a telecommunications license or to speed up approval. But doing so is against U.S. law. "American companies, joint ventures and Americans doing business in Russia should realize that [under the Foreign Corrupt Practices Act passed in the late 1970s] it's against U.S. law to engage in any corrupt business practices," said Burt. But Jonathan Russin of international legal counselors Russin & Vecchi LLP (Washington) points out that it's highly unlikely telecommunications companies in Russia will have to make low-level bribes or "grease payments" to connect with government officials on things like telecom licenses. (c) CMP Media, Inc |
© Copyright 1995-2000, The Motley Fool. All rights reserved. This material is for personal use only. Republication and redissemination, including posting to news groups, is expressly prohibited without the prior written consent of The Motley Fool. The Motley Fool is a registered trademark and the "Fool" logo is a trademark of The Motley Fool, Inc. Contact Us