Low-k stalls for a while

The low-k dielectrics market may be slow getting out of the starting gate, according to a recently completed study by Kline & Company, but the long-term potential for the technology sector remains strong.

The low-k dielectrics market may be slow getting out of the starting gate, according to a recently completed study by Kline & Company, but the long-term potential for this technology sector remains strong.

The market for low-k dielectrics – materials exhibiting a low dielectric constant that are needed for leading-edge semiconductor fabrication – is projected to expand from only about $16 million this year to nearly $400 million by 2006.

‘The real ramp-up should occur in 2003 and 2004, when production at 100-nm design rules start coming to commercialization,’ says John Davis, business manager at Kline & Company. The market will rise by two to four times the previous year’s level in each of these two years, according to Davis.

In 2000, the dielectrics market was expected to start accelerating significantly by 2001/2002. However, that simply didn’t happen. According to the new study, the two-year delay was due to a number of factors, but chief among these were the fact that fabricators delayed implementation out of fear that low-k dielectrics were unreliable and were unlikely to be fabricated at high yield.

They worked around the problem by redesigning circuits to minimise resistance-capacitance (RC) delays, by extending fluorinated silicate glass (FSG) technology into the 130-nm design node, by introducing trimethylsilane for the 130-nm node, and by emphasising copper over novel dielectrics as the integration route of choice.

On the web