Moore’s Law to die at 18 nm, analysts predict

Posted

El Segundo (CA) – iSuppli most certainly will initiate yet another discussion whether Moore’s Law, a forecast made by Intel co-founder Gordon Moore that the number of transistors on an integrated circuit will double every 18 – 24 months, can be upheld in the future. The market research firm believes chip makers will crash into financial barriers as soon as the 20 nm mark is passed.

We have heard it before and we will hear it in the future again. The rules of Moore’s Law, one of the most significant guidelines for chip makers and semiconductor innovation, cannot be sustained anymore. But we notice that the frequency of this prediction is increasing and the reasoning behind such claims is shifting. It is not so much that the transistor count can be increased through shrinking structures, it is the production cost of smaller transistors that may break Moore’s Law’s neck.

“The usable limit for semiconductor process technology will be reached when chip process geometries shrink to be smaller than 20 nm, to 18 nm nodes,” said Len Jelinek, director and chief analyst, semiconductor manufacturing, for iSuppli. “At those nodes, the industry will start getting to the point where semiconductor manufacturing tools are too expensive to depreciate with volume production, i.e., their costs will be so high, that the value of their lifetime productivity can never justify it.”  

In the end, iSuppli believes that Moore’s Law will lose the perception of the ultimate chipmakers’ guideline it has today. By 2014, “Moore’s Law will no longer drive volume semiconductor production,” iSuppli said.

We are pretty sure that Intel and other chip makers will deny that Moore’s Law’s death can be predicted, especially since Gordon Moore said back in 2007 that his prediction can be upheld for at least another 10 years. Also, Toshiba yesterday unveiled plans for 16 nm structures.

Of course, the question is how companies can mass-produce such small structures and whether they can afford it. Upcoming 32 nm production facilities already cost more than $4 billion and only make sense for chip makers with revenues of more than $10 billion, at current financial depreciation models. The dramatic increase in cost drives joint-ventures between chip makers and allows only companies such as Intel and Samsung as well as contract chip makers such as Globalfoundries, Chartered or TSMC to operate their own fabs.      

Eventually, Moore’s Law will run into the limits of nature. The question is when that will be.