 |
 |
June 2003 |
Silicon Valley BuzzSupported
by The Hoffman Agency China U.S. still vulnerable to cyber attack More than 20 months after the Sept. 11 terrorist attacks, the United States remains ill-prepared to defend against a strike on the nation's critical computer systems because of slow-moving federal research efforts, members of Congress said recently. They charged that instead of working at breakneck ``Internet time,'' the four key agencies charged with researching new technologies to combat cyber attacks are stuck in the glacial world of ``government time,'' still crafting memorandums of understanding to allow collaboration on projects. While defending their efforts and saying progress was being made, the agency heads acknowledged there is much more work to be done. Terrorism experts fear attacks on computer systems that operate electricity grids, phone systems or other critical infrastructure as part of a terrorist strike. The federal government, in conjunction with private industry, has been trying to protect those systems through the use of fire walls and other technology to prevent such attacks or lessen their impact. The vulnerability of a cyber attack is particularly acute for the U.S. military, which is becoming increasingly dependent on computer networks and information technology, said Tony Tether, the director of the Pentagon's Defense Advanced Research Projects Agency, or DARPA. The testimony follows the departure of two key White House cyber-security advisers earlier this year. The upheaval has led to concern in the high-tech industry that the Bush administration is not making cyber security a priority in combating terrorism. Sharing those concerns, Congress last fall passed the ``Cyber Security Research and Development Act,'' which authorized $903 million for research efforts over the next five years. In creating the new Department of Homeland Security, Congress set up a Science and Technology Directorate to oversee cyber security as well as other uses of technology in counterterrorism. The heads of the four lead agencies for cyber-security research -- the directors of the science foundation, DARPA, and the National Institute of Standards and Technology, along with the undersecretary for science and technology at the Department of Homeland Security -- said they were making progress and beginning to work collaboratively on projects. But some science committee members were critical of their efforts. Tether complained that DARPA had money to spend on cyber-security research but lacked proposals, while Colwell said her agency had too many proposals and not enough money to fund them. Boehlert also criticized the agencies for not putting more resources into cyber research. For example, the Department of Homeland Security's science and technology division has requested $803 million in its 2004 budget, but only $7 million is earmarked for cyber-security research. Last fall's legislation authorized the National Science Foundation to spend $110.25 million on cyber-security research, but the agency is requesting only about $51 million. DARPA's unclassified budget for cyber-security research has actually declined, from about $90 million in 2000 to $30 million in 2003. But Tether said those figures were misleading, because more projects are now classified. He estimated the agency will spend about $100 million on cyber-security research in 2004.
Silicon Valley had hoped to break out of its malaise with a sharp rebound during the past three years, but the valley and its largest public companies are still waiting for the long-anticipated turnaround. The largest 150 companies in Silicon Valley lost billions of dollars as sales of technology equipment fell for an unprecedented two years in a row. The annual Mercury News listing of Silicon Valley's biggest 150 public companies shows that during the latest four quarters, the 150 companies lost $18.4 billion on $265.4 billion in sales. A year earlier, the SV150 lost $89.8 billion on $251.5 billion in sales. This means that technology customers that bought the valley's Internet equipment, software, computers and chips got a bargain in 2002. They paid $93.07 for goods and services that cost $100 to produce. For a valley accustomed to spurts of hyper-growth, the past three years have been a wrenching change. Forecasts among market researchers call for minimal growth in information tech spending, ranging from 3 percent to 5 percent this year. That's well below the historic average. If the tech economy were to rebound quickly, a surge in sales would boost profits quickly because costs have been cut so sharply. But that doesn't change one key problem of this new, more sober world: There are still too many tech companies chasing too few customers. More than half of the SV150 companies -- 79 in all -- are losing money, and many are burning through cash to stay in business. At the same time, in true Darwinian fashion, the strongest companies -- particularly Intel and Cisco Systems -- are getting stronger, while the weak are falling behind. In Silicon Valley, it has also become unusually difficult to know exactly what has happened to the largest 150 companies. Peel back some structural change, such as large mergers and large write-offs, and the picture is actually worse: Two big mergers skew 2002 sales for the largest companies. On the surface, it looks as though this year's SV150 increased their sales by less than 1 percent in 2002 -- the year that Hewlett-Packard succeeded in merging with Compaq; and Sanmina combined results with SCI. However, if these two newly merged companies are dropped from the SV150, the remaining 148 companies saw their sales shrink 4 percent last year. That's the second year of shrinking sales for Silicon Valley companies, something that has never happened during the nearly 15 years of keeping the statistics. The profit picture is fuzzy. For one thing, Intel and Cisco together account for $6.2 billion in net profits for 2002, offsetting a quarter of the losses of the remaining 148 of the SV companies. For another, three companies -- Sun Microsystems, Solectron and Sanmina-SCI -- wrote off $7.9 billion as a group, accounting for almost half of the total lost by the SV150 in 2002. Sales fell more sharply for the smaller companies that the larger group. In fact, the largest 74 companies saw sales fall by 3.8 percent from their comparable period last year, but the smaller 74 companies saw their total sales fall by 6.5 percent. Profits also were worse: The top 74 companies had red ink on the bottom line, losing about $5.55 for each $100 of products sales. But the bottom 75 gushed red ink, losing $41.25 on every $100 sold. And investors dumped the stocks of the SV150. The SV150's market capitalization fell 35 percent to $601.1 billion from March 2002 to March 2003 -- worse than the 28 percent decline in the Nasdaq composite index during that period. For overall information technology, IDC expects spending will rise 2.3 percent in 2003 and then 6 percent the next year. The best growth will come in software and services and smart handheld devices. The worst market will be mid-range computer servers and traditional workstations.
Despite two months of intensified lobbying, the technology industry has made little progress in changing minds and slowing the momentum to make companies count their stock options as a business expense. Since some of Silicon Valley's heaviest hitters lobbied Congress and the White House in March, supportive politicians have introduced bills to derail the proposal. Corporate leaders have warned that counting stock options as an expense would turn profits into losses at many tech companies. But no hearings are scheduled and there are widespread doubts that Congress has the will -- as it did in the early 1990s -- to force the accounting rule-makers to back off. And though a round-table discussion on Capitol Hill recently allowed proposal opponents to rail at the chief rule-maker, no minds seemed to have been swayed. Indeed, two powerful players -- Federal Reserve Chairman Alan Greenspan and Securities and Exchange Commission chief William Donaldson -- later reiterated their support for expensing. Perhaps most important, none of the jousting seems to have deterred the rule-making board, known as FASB, from its plan to issue a proposal to count options as an expense in the fourth quarter -- and implement it by April. Don't expect the tech lobby to give up the fight, however. Indeed, recent events highlight the battle plan that the anti-expensing forces will use in the coming months: Stall for time. It's hard to imagine when conditions could be worse for battling against expensing stock options. The tech slump has weakened the industry's political clout. Stock options are now linked in the public's mind with executive excess and corporate scandal. And Congress is preoccupied with Iraq, homeland security and the economy. The foundation of the effort to buy time are two similar bills introduced in the House and Senate last month that would force the SEC to study the impact of any FASB rules for three years before implementing them. No hearings have been scheduled. Cast doubt on the numbers. The tech industry's leading argument is there's no reliable way to estimate the cost of stock options -- and that such estimates will distort earnings and confuse investors. Turn expensing into an economic issue. So far, the debate is in the hands of FASB's accounting experts, but reframing the issue as one with economic implications opens the door for Congress to weigh in. The tech lobby warns that the impact of expensing would ripple through the economy. Tech companies that dole out options would see their earnings and stock prices plummet. Cash-strapped start-ups no longer would ladle out options to woo workers. And expensing would give Asian competitors that embrace stock options an advantage. Drive home the impact on rank-and-file employees. Much of the recent backlash against options was triggered by executives who pocketed millions in stock-option profits while running companies like Enron and WorldCom into the ground. The tech lobby's goal is to show that expensing could further slow the flow of options to workers.
Wireless home computer networking is becoming more and more popular, even as the buying process gets more and more complicated. Just when consumers were getting comfortable last year with the concept of 802.11b, also known as WiFi, the industry trotted out the 802.11a standard. When 802.11a turned out to be a flop, the networking industry rolled out 802.11g. The story gets even more confusing with the arrival of A+G networking gear that uses both the 802.11a and 802.11g standards. To begin at the beginning: Computer networks are almost mandatory for homes that have a high-speed ``broadband'' Internet connection, such as cable modem or DSL phone line, and want to share that connection among two or more computers. The cheapest way to build a home network is with Ethernet cables, which look something like fat phone wires. For less than $50, you can get a good Ethernet router -- a kind of switch box which shares the cable modem or DSL connection among multiple lines. But many people
can't or won't run Ethernet cables throughout the house to get access
in every room. Routers with built-in 802.11b wireless networking now cost well under $100, while 802.11b PC Cards for notebook computers and USB adapters for desktop computers cost under $50. There's no need to run cables around the house, and you can work online with a laptop anywhere within range of the router -- even the backyard or the porch. Hardly anyone needs more than 802.11b at home. Operating on the 2.4 gigahertz (GHz) spectrum, shared with other devices including cordless phones and microwave ovens, 802.11b has a peak speed of 11 megabits (mbps) per second. In the real world, speeds range from 2 to 5 mbps -- still more than enough to share home broadband, which typically doesn't exceed 1 mbps. And 802.11b does well at other common networking tasks, such as sharing printers and files or setting up multiplayer computer games. Just about the only things that exceed 802.11b's capacity are moving huge files -- in the hundreds of megabytes -- or sharing high-quality video. But 802.11b
is already starting a slow fade and will mostly disappear late this
year or early next year.
|
