SIGNS OF A ROBUST RECOVERY FOR SEMICONDUCTORS abound. Intel shipped a record number of microprocessors last year, demonstrating continued demand for PCs, servers, and laptops. In an odd coda, revenues for the company only grew a scant 1 percent because of intense price competition, but low prices drove volume, particularly in Asia, where companies are racing to build IT infrastructure.

There is also the unrealized promise of the Internet for communications, logistical, and e-commerce applications. Internet traffic doubles annually, and the sales of goods and services are growing at rates much faster than their traditional counterparts. There has been a huge investment made in equipment and fiber-optic transmission lines to handle all this traffic, but it remains under-utilized, creating a tremendous opportunity for more people, with more hardware, to jump aboard.

Major IT investments have doubled productivity for American business since 1995, and most industries recognize the need to continue spending, viewing it as less a way to save money, and more a path to increased profits. There are also huge sectors of the economy – healthcare for instance at 15 percent of U.S. GDP – that have far to go in automating their systems. Government represents another ripe opportunity for modernization, the FBI is just in the process today of equipping all of its agents with e-mail. So spending for business will continue for competitive reasons, and spending in other sectors will grow rapidly because consumers will demand improved service for their dollars.

And speaking of consumers, think of our world, and how everything connecting or entertaining us now is digital – entertainment, photography, communications – this is an unstoppable megatrend that is fueled by silicon advances. Even more important is the fact that much of the planet has yet to visit the Internet. Only 12 percent of citizens worldwide own a PC and are online. As prices continue to decline, standards of living rise, and economies of scale advance globally, these underserved consumers will gobble up as much product as the industry can ship.

Over the next few years I see tremendous growth across the entire electronics spectrum, and a few real standouts with tremendous mass-market appeal.

Mobile computing is everywhere. In May of this year the sales of laptops (in dollars) surpassed desktops for the first time. Many users are foregoing the desktop altogether, as new models have the memory, speed, screen size (up to 17 inches!), and power to do any job. Even more interesting is the widespread adoption of mobile e-mail and messaging, exemplified by standout products like RMI’s Blackberry and the latest “smartphones” for the handset leaders. European analysts expect 40 percent of users to send and receive e-mail from their smartphones by 2008, up from just 1 percent today.

These mobile platforms are driving myriad silicon projects, including low-power microprocessors, denser memories, flash, I/O components, mixed signal designs and wireless connectivity.

Wireless is a good example of the type of opportunity that typifies great silicon success stories of the past. As mobile computing took off, users sought ways to become untethered from Ethernet, dial-up, and broadband. The 802.11 wireless LAN standard quickly matched, then surpassed, Ethernet speeds, and can now deliver an astounding 54Mbps. The market will more than double to just over $3 billion in four years, and unit shipments of 802.11 silicon will rise 60 percent this year alone.

Taking off in other ways as well, these wireless networks may soon replace the rat’s nest of cables behind the family entertainment center, while simultaneously serving as the home network of choice. Even more alluring, wireless offers a terrific “last-mile” option for satellite and cable providers wanting to wire the home for big-pipe, broadband services.

And, as a personal aside, those cables can’t be gone too soon. The digital living room has arrived, and sales of products to amuse the masses are being snapped up at a record pace, but who wants to hook them all together? Give me the wireless option now!

Television continues to be the delivery vehicle of choice, but it is evolving rapidly. Household HDTV sales will soar past $15 billion this year according to the Consumer Electronics Association, accompanied by similar commercial gains for transmission equipment, cameras, studio gear, and related items.

Digital recording and playback is also in vogue, as the spectacular penetration of DVD players attests. DVD sales skyrocketed twenty-fold in the past five years as prices plunged from $400 on average to less than $100 today. Right on its heels is the emerging PVR (personal video recorder) category, with the best-known brand TiVo. Although unit sales just broke through the 1 million per year mark, this is expected to gain user acceptance just as rapidly as DVDs in the next few years since an overwhelming majority of PVR users (topping 80 percent) report they are “extremely satisfied” with the hassle-free technology for recording and playing back television programs.

Last but not least are the gaming boxes, with Sony’s PlayStation 2 already surpassing unit sales of 50 million since its 2000 rollout. Followers Nintendo and Microsoft will have sold another 30 million or so by year’s end.

All of these digital entertainment options will require vast amounts of silicon–some highly specialized, some commodities –to produce the images people are clamoring for. Companies like NVIDIA are reaping the whirlwind with advanced graphics processors, and drive makers and memory suppliers are encountering vast new markets as digital recording supplants the VCR.

Another wonderful opportunity is digital photography. Just this week I viewed an online ad for a $99 camera/photo printer combo from HP – definitely in the mass-market’s sweet spot. Little wonder then that sales rose almost 30 percent in Q2 of this year to slightly below 10 million units. The opportunity for CMOS sensors seems unlimited here, as cameras continue to morph into lots of applications for security, in cell phones, desktop videoconferencing, and more.

All of these things bode very well for the semiconductor market. The only cloud on the horizon, in the midst of all this good news, is that people are still a bit nervous about taking advantage of the opportunities presented, and overcoming a series of increasingly difficult design and manufacturing challenges. As Andy Grove said recently: “Technology is unstoppable – but it’s delayable.”

The impact on startups has become extreme: this year, just 15 percent of VC dollars are going to new startups, a decline of 20 percent since 1995. Correcting this situation to tap the innovative spirit of young companies will hasten the introduction of new technologies, once more growing the pie for everyone.

On the technology front, the ITRS roadmap is predicting a slowing process migration from every two years to every three years. There is also concern that CMOS scaling may be nearing its end, and designers and EDA tool suppliers will be confronting a number of issues with no known solution as line widths shrink below 90 nanometers. Chief among them are the control of leakage power, consistent transistor performance, increased mask costs, process variation control, and test issues. The roadmap foresees significant design and architectural changes in the years ahead that will exacerbate these problems.

For instance, supply voltages will slide from 1 Volt today to 0.7 Volt by 2007. As these voltages drop in nanoscale silicon transistors can unexpectedly perform faster than designed, and generated heat can zoom off the charts. There will also be difficult timing-closure issues as frequencies more than double to nearly 7GHz in high-performance chips. Unless designers are examining these effects with a complete, concurrent view of heat, timing, power, signal-integrity, and electromigration, they will have an extremely difficult time getting to silicon. This is due largely to the fact that each of these phenomenon begin to have new and devastating effects, both alone and interactively, as designs plunge below 90 nanometers.

Solving the money issue will be easy – the demand is there, the ideas are there, so the dollars will follow. Technology hurdles will fall at the same time, as the entrepreneurial engine once more drives the business to new levels of innovation and productivity.

Vic Kulkarni is the president and CEO of Sequence Design. Prior to joining Sequence Design, Kulkarni was the General Manager of Avant!'s Silicon Business Unit, managing the R&D and Product Marketing functions worldwide. Additionally, he held the position of Vice President of Marketing at Avant! and was involved in setting the company's intellectual property (IP) strategy. He holds an M.S.E.E. in Solid State Electronics from University of Cincinnati, Ohio, and a B. Tech. from the Indian Institute of Technology, Bombay.