In a beauty contest of Asian economies, there is probably nothing wilder or more attractive than the rough-and-tumble marketplace in China. Powered by annual GDP growth rates exceeding 7 percent, hundreds of millions of urbanites are entering a new phase of consumerism that mesmerizes foreign companies.

At the moment, the hope and hype assigned to China means it would take a disaster of monumental proportion to stall progress. China has too many smart people to fail, too much cheap labor to be uncompetitive, and a lot of ambition to see it through. The Technology Revolution is just taking root in China, in very fertile soil. Somewhere in the distant future, far, far away, China will become an influential player in semiconductor design and manufacturing.

What's likely to happen over the next three to five years or so is this. China will continue to focus on its chip industry as an engine of modernization, dangling the ever-present carrot of its domestic consumers as a lure for foreign investment.

The country's economy will grow, more people will purchase PCs and cell phones, its labor pool will remain relatively cheap and seemingly inexhaustible, meaning it will snap up more skilled tech jobs that might have gone to American or European engineers, and unskilled assembly-type jobs that could put Mexicans and Eastern Europeans out of work. In short, it will exploit the basic trends of globalization.

But, today, next year and probably for the next decade, the hype is just background noise. Smart money will stick with Taiwan and South Korea, two key hubs of IC design, system integration and manufacturing that will not slip away into irrelevance.

In China, true innovation at the chip level is still hard to come by, and will be for some time. Many of China's 400-plus design houses are mom-and-pop shops, which brought in $150 million in 2001 and $157 million in 2002, said market research firm iSuppli. In global terms, that represented about 1 percent of revenue for the fabless sector in those years.

As demand ramps, China's fabless IC design industry is expected to surge to more than $800 million by 2006. Still, Taiwan's top fabless company, Mediatek, will take in that much this year.

Yet, there's no question major demand exists. Market researcher IDC expects China's PC sales to nearly double in a few years, from 11.3 million in 2002 to 21.1 million in 2006. From 2001 to 2005, handset sales should rise from 44 million to 70 million, IDC said, a 59 percent jump. It's worth noting, however, that competition for both foreign and domestic vendors will be among the most intense in these product areas.

During the next few years, IT companies will need to ward off another possible export from China: deflation. In chip-making, for example, China is set to ramp 13 more 8-inch wafer fabs by late 2005, iSuppli reports, with at least 10 seeing volume production by year's end.

Some other things to keep an eye on will be China's progress on IP protection; job development for the more than 150 million unemployed-a potential source of civil unrest; pervasive corruption; a shaky financial system; and the privatization of a mammoth network of often-inefficient, unprofitable state-owned enterprises. This is by no means a complete list of hot-button issues, either.

Here and there, some neat technology developments are happening in China. At the Chinese Academy of Sciences, researchers are working on grid computing projects. In Beijing, a small CAS-spin-off called BLX IC Design Corp is trying to win sockets for a nearly MIPS-compatible 32-bit embedded processor while it works to field a 64-bit version.

But much of the advanced work in China is academic, and slow to transition into commercial application. "We are trying to find a better way to do our research," said Xu Zhiwei, a deputy director at CAS' Institute for Computing Technology. "We know it is not as efficient as our colleagues in the United States or Europe. So we are trying to learn from them."

To find the most interesting, and truly innovative, development this year in China, one would have to hit Hong Kong, the more tech-savvy stepchild of China. In what at first seems laughable, Culturecom, an old kung fu comic book company, has launched an IBM PowerPC-based system-on-chip equipped with a unique Chinese character-generating algorithm.

The algorithm helps the chip generate more than 32,000 characters, in just 256 kbytes of on-die memory, handily beating its competitors. With strong ties to the Chinese government and a preference for Linux-based systems, Culturecom could become a key player in embedded systems aimed at the Chinese market.

Beyond that, the hunt for innovation in China turns into a bit of a sojourn. A quest for the patient, inquisitive soul.

A Western engineer's worst nightmare has got to be a guy in a far-off place who works longer and perhaps harder for less money, yet produces a product with acceptable quality. That is Taiwan in a nutshell. (And in the far-off future, beyond 10 years, that will start to be China, too.)

In Taiwan, companies in the IC design world won't always settle for being second-best. The products will start to be more sophisticated and they will cut down the time between when U.S. suppliers release a product and they do, shortening the space of time U.S. companies enjoy high margins.

This is already happening. For starters, just look at the markets for DVD and RF chips. In the past few years, Mediatek Inc. and ALi Corp. have become important suppliers to Chinese makers of DVD players. More recent has been the rise of a few small analog design companies, such as Ralink and Airoha, which are targeting the wireless LAN market.

Recently, there have been some interesting indicators of Taiwan's ability to move up-market. For example, Sunplus Technology is transitioning from an IC maker for low-end toys to a designer of high-end DVD and DSC chips. To speed its development in those new areas, the company has forged alliances with Oak Technology and Silicon Image.

This could be the future model for Taiwan-Silicon Valley relations, said Wayne Shen, an executive at Sunplus. "Traditionally, Silicon Valley companies only enjoy the growth period of the product cycle, but they can't enjoy the volume as much because of greater competition," Shen said. "Now, with back-end help, they can get good margins in both."

There have been other partnerships, too. Xtreme Graphics Innovation (XGI) is a startup so new that the smell of construction glue still overwhelms its fancy digs in Taipei's high-end, high-tech Neihu District. Despite its small size, the chip designer is determined to survive and thrive in the fast-changing, intensely competitive graphics market.

Xtreme is getting help from the graphics division of Trident Microsystems, which Xtreme recently acquired. XGI will combine those resources with the former graphics division of Silicon Integrated Systems, one of its major investors, to create products for the mid and high ends of the market, where they will bump heads with the likes of Nvidia Corp. and ATI Technologies.

In the crowded wireless LAN arena, another startup, called ZyDAS Corp., is highlighting a proprietary multipath decoding technology that it says improves data throughput by more accurately interpreting wayward signals. ZyDAS is also pushing an overdrive function that pushes 802.11b to 16 Mbits/s when paired with another ZyDAS chip.

This is what the Taiwanese will remain good at for many years to come: Taking a technology and making it better and cheaper. They may not be chasing the newest technology, notes Norman Hung, president of ZyDAS, but when a market really starts to look feasible, then Taiwan firms are ready to pounce.

Hung, like many Taiwan executives, doesn't see China as an immediate threat. If anything, the Chinese will focus on low-end chips for consumer white-box goods, such as refrigerators, air conditioners and TVs, while Taiwan continues to move up market.

"Just like we took that low-end business from Japan 10 years ago, China will take it from us. It is low-barrier stuff. But in higher integration and higher value products, we still have a good chance to compete," he said.

Hynix, for better or worse, will hang on. The government just isn't willing to let it fail. Despite tariffs in the U.S. and Europe, the company lingers and insists it will soldier on.

Korea is more well-known for its chaebols, meaning that plenty of small companies are starved for attention. But small firms keep emerging in Korea to prove that the country is more than Samsung, Hynix and LG. The problem is reaching a critical mass of fabless firms that will help to diversify the backbone of Korea's IT industry.

Many of Korea's engineers prefer the big firms, especially after Korea experienced its own Internet bust not too long ago. Some engineers had taken chances on small firms at the time and got burned. Now that some of them have returned to the culture of the chaebol, it may be tough to draw them out again.

Or maybe not. During the past few years, a whole new breed of small handset design firms has cropped up in South Korea, many of which sell their designs on to Chinese manufacturers. Other firms are focusing on games that can be played interactively among groups of players along disparate points of Korea's high-speed broadband networks.

In semiconductors, a few small companies are starting to catch some attention, too. With its compact cell size, Silicon7 Inc. is pushing aggressively to win space for its pseudo-SRAMs in 2.5 and 3G cell phones. And low-cost RF-CMOS technology is helping GCT Semiconductor Inc. build a reputation as a serious threat to the usual suppliers of RF front ends in wireless LAN chip sets.

Despite those examples, the fabless activity is small in Korea and seems set to remain that way for some time. Observers in Korea note that the fabless industry is seeing more activity, but it is small when compared with Taiwan or the U.S.

In the long run, this could weaken the sector by maintaining its bet on the big companies to always see it through. The trials of Hynix Semiconductor should serve as a wake-up call to industry and government officials that the Korean machine needs to keep diversifying. "Considering all the Korean businesses, in semiconductor, LCD and handsets, there is no good, sizable fabless semiconductor company," said KH Lee, the president of GCT.

He points to Taiwan as an example for Korea's industry, noting that the fabless industry there had good support from the government and is still in a good position to compete for some time to come, despite the rise of China. "There are many good, financially healthy medium sized companies in Taiwan. In order for Korea's economy to be more healthy, we need to find a Taiwan style," he said.

That assignment will fall on the shoulders of a small but growing number of Korean entrepreneurs, who may appreciate the chaebols for their vast resources but ultimately shy away from their cumbersome nature.

Mixing a bit of Taiwan's nimbleness with Korea's strong sense of entrepreneurship could yield an even more powerful regional competitor. If Korea maintains its more monolithic IT industry, where tech giants are the norm, then it could well follow Japan into its own Lost Decade.

The trick for these emerging companies will be to leverage Korea's strengths, such as memory and cell phone design, to enter new or existing markets and move up the value chain.

"I don't see Korea as falling into the same trap as Japan," said Woodward Yang, CEO of Silicon7. "Koreans are well aware of the way that Japan has gone in the last 10 years, and they are trying to avoid that."

To date, the edge of the Western-trained engineer is that he likely works smarter, i.e., he is more creative in determining what people want and getting it to them first. This is an intangible, yet priceless, asset, for it may perpetually help Western companies in their desire to stay ahead.

But in the semi industry to date, much of the demand has come from Western countries and their powerhouse economies. Over the next several years, that won't change. Beyond that, though, the Asian nations will chew up a greater piece of global demand, giving their domestic firms an opportunity to figure out what they want. Just look to Japan as a leading-edge example in Asia.

This doesn't mean foreign firms won't be able to cater to Asian preferences, especially if they continue to employ locals and listen to their feedback. It's just that the market will be more competitive, as it should be. And more jobs will go to locals, which seems natural since the growth is occurring in their markets.