Chinese start-up has ambitions in a market overlooked by US and Europe

Michael Logan in Shanghai

A Shanghai start-up hopes to become one of the mainland's biggest chip-design companies by focusing on a niche its United States and European peers have ignored: radio frequency chips for xiaolingtong handsets.

Comlent Technology is targeting revenue of more than US$50 million annually by 2006 - in a market where the top five mainland chip designers collected no more than $25 million last year.

The mainland company believes it can reach this goal by taking advantage of changes in the handset manufacturing industry. Its approach is to combine as many as six radio frequency chips into one or two chips, thus reducing a handset's overall cost.

This "cost-down" model has served Taiwan's chip designers well for many years, helping the industry achieve an estimated US$5 billion in revenue this year.

The US and European companies which dominate the market for radio frequency chips have overlooked xiaolingtong because it is used mainly on the mainland and in Japan, where it is known as the personal handyphone system (PHS).

However, their unfamiliarity with the niche service has left an opening for Comlent.

"The domestic market is open to us. We're careful in positioning our products," Comlent chief executive Kai Chen said. "US and European companies do not care [about xiaolingtong]."

Mr Chen said the company's chips could reduce the cost of a xiaolingtong handset from US$30 to $25. For cordless phones, which also use radio frequency chips, the cost reduction is about $3 to $22.

"That 10 per cent is still very significant, very meaningful for Chinese [original equipment manufacturers]," he said.

Mr Chen estimated mainland manufacturers shipped 60 million cordless phones last year and had 90 per cent of the global market.

Comlent is confident volume production will begin by the middle of next year. Its foundry partners are Taiwan Semiconductor Manufacturing Co, which produces chips for Comlent using 0.25-micron processing technology, and Jazz Semiconductor of the US, which is making 0.35-micron chips.

Mr Chen said Comlent was likely to move some production to mainland foundries within two years. "All the [mainland] foundries have approached us," he said.

The company expects as much as US$10 million in revenue next year, $35 million in 2005 and more than $50 million in 2006. A profit is expected in 2005 and gross margins are forecast to be 50 per cent. A Nasdaq initial public offering would follow in 2006 or 2007.

All this sounds ambitious for a company which has yet to ship any product, but the start-up's business plan has found support in the likes of Tsai Ming-kai, chairman of MediaTek, the world's largest maker of chips for DVD players.

Mr Tsai was one of several investors who put a combined US$2 million into Comlent's first round of venture capital fund raising. A second round is in the offing and is expected to raise US$5 million.

A potential second-round investor said her firm liked Comlent because it was a well-managed company run by a group of returning overseas Chinese.

"They have this opportunity which is a very good market niche," she said. "PHS is a China phenomenon and has a potentially larger market in other developing countries."

Mr Chen said one advantage his company would enjoy was a proximity to mainland cordless phone and handset manufacturers, allowing Comlent to better serve them.

Too often overseas companies design chips and then expect their customers to design products to use those chips. "The Chinese manufacturers try to build systems around that chip - good or bad," he said.