In 1958, the new Texas Instruments Inc. lab on the outskirts of Dallas had emptied out for summer vacations. Jack S. Kilby, a new hire, stayed behind and worked on some sketches. He was a 34-year old engineer who'd been fascinated with electronics since high school, when a storm knocked down telephone lines in Western Kansas and forced his father to use ham radio to communicate with far-flung employees of his power company.

That summer in 1958, Kilby drew a sketch of an IC that crammed the essential electronics for a semiconductor on a sliver of material half the size of a paper clip. Instead of using discrete components, wired or soldered together, Kilby's glossy circuit was built in layers on a single substrate or wafer of germanium.

Fast-forward 50 years, riding on a remarkable growth trajectory known as Moore's Law, the IC industry seems to be facing an interesting crossroads. While the relentless pursuit of scaling IC to smaller geometries continue, observers are suggesting that time has come to address both technical and economic limitations of Moore's Law.

State of technology
Chipmakers are rolling out 45nm ICs; 32nm designs are in the works; and R&D has started for 22nm and smaller nodes. But experts say that the delivery of chips at 32nm and beyond will face some critical challenges, which may make or break the future of IC scaling. Challenges, however, are nothing new in the industry's extraordinary journey. In fact, challenges have defined this hyper-competitive business decade after decade.

Much of the 1960s was spent in shaping the IC manufacturing landscape. Silicon Valley began to rise as the center of the IC world. We had landmark innovations like the advent of DRAM.

While the development of the microprocessor was the highlight of the 1970s, the next decade saw the inception of ASIC and FPGA and the birth of the fabless model. IC saw its shining moments in the 1990s as it accomplished new levels of excellence in design and manufacturing and drove high-volume products in mobile, PC and Internet markets.

And here we are in 2008 debating the future of Moore's Law. To mark the 50th anniversary of the IC, TI announced Kilby Labs on Sept. 12, exactly 50 years after Kilby demonstrated the first IC in a prototype phase shift oscillator. Kilby Labs will bring together university researchers and TI engineers to seek new opportunities in semiconductor technologies. The inception of Kilby Labs is a testament that the spirit of innovation is alive and well in semiconductors, one process node at a time.

Back to top

• Opinions Archive