G. Pascal Zachary writes,
Designers led by Mr. Bohr in Hillsboro, Ore., chose hafnium to replace silicon oxide, the venerable insulator in chips and a material used in making glass. Mr. Bohr also helped to identify new materials, whose identity Intel is keeping secret, for the crucial transistor “gates” that sit atop a chip’s insulators.
On Nov. 12, Intel will begin shipping its first chips using the new processes. Gordon E. Moore, Intel’s co-founder, recently declared that the hafnium-and-gate process innovations should allow his so-called Moore’s Law, whereby chips grow ever faster and less expensive, to hold true for some time.
Despite the enormity of the achievement, Mr. Bohr is relatively anonymous, even within Intel. “The work of process development comes second to creating new designs for chips,” he says. Not surprisingly, when Intel starts shipping the new chips, neither the hafnium nor the gates innovations will be trumpeted as selling points. Rather, Intel will emphasize how customers can benefit from using the chips.
The point of the article is that the processes--what I like to call the recipes--that enable new products and services are often hidden from the public view. In general, as the information content of goods and services rises--and Ray Kurzweil has long argued that the information content will asymptotically reach 100 percent--the sources of economic success become less visible and less tangible.
Thanks to Nick Schulz for the pointer.
"Ray Kurzweil has long argued that the information content will asymptotically reach 100 percent"
Too utopian an idea if you ask me. Any process needs a low-entropy source at the input, producing much higher entropy at the output.
the sources of economic success become less visible and less tangible.
On a somewhat related note, productivity ensures that there are less and less people needed to do any one job. Thus, for the average person, the personal link to any line of work is more than likely not there, and becoming even less so.
We're getting to a situation not unlike that classic Trek episode with the planet where the computers run everyhting and the inhabitants don't understand how anything really gets done.
It's a nice place to live, except when the computer breaks!
Not true, at least among engineers.
Buzzcut: Not to worry! Everything will be fine until the computers build themselves -- then it'll get hard to fix things when they break. But by then everything will be so cheap that no one will care when something breaks.