Can You Work on a Computer That Won't Be Built for 10 Years?

Think about it: In the past few decades, computers have gotten increasingly faster, smaller, and more powerful — at an incredible rate.

Back in the 1960s, Intel’s co-founder Gordon Moore predicted that chip performance would double every 18 months (and the cost would come down). Ever since, the so-called “Moore’s Law” has largely held true — until now, when the industry is approaching a critical microchip roadblock. Which is rather unsettling to quite a few people in this $200 billion global industry.

Aluminum Foiled
The primary challenge is not the transistors. We can now make these at microscopic size and pack millions onto a single chip. It’s the interconnect, the wiring or pathways that link the transistors. The industry’s current interconnect material, aluminum, is limited by its electrical resistance and reliability. We appear to have maxed out its conducting abilities — which will soon cap the peak performance of the chips themselves. It’s time for a new solution that will drive the better, faster computers of the new millennium.

Breakthrough Research at Rensselaer
The university has spent the past decade focusing on interconnect issues — tackling the problem before many in the semiconductor industry were taking the crisis seriously. Rensselaer researchers have performed pioneering work in areas such as silicides, chemical mechanical polishing, and low dielectric constant polymer insulators. (All of which have been adopted or are under consideration throughout the industry.) Perhaps Rensselaer’s most daring research goal, however, has been the use of copper as an interconnect.

“A Copper Interconnect? You Must Be Joking?”
In the not-too-distant past, many in the semiconductor industry scoffed at the notion of a copper interconnect. But in September 1997, IBM announced that it was going to use copper wiring in its new, improved chips. IBM’s stock price jumped, competitors gasped and pointed to their own copper research, and journalists called IBM for an expert explanation. The company referred them to Murarka at Rensselaer as one of the best sources of information about the new technology. And now copper research at Rensselaer is moving into more reliable and simpler to fabricate copper-alloy technologies necessary for sub-tenth micron technology.

Then Top It Off With Teflon
Copper is, indeed, a far better conductor than aluminum and appears to be the interconnect solution for at least the next decade. However, it has to be surrounded with a insulator material, which also provides reliability. As it turns out, Teflon may be that insulating material, but it has to be vapor-deposited to be applied (we’re talking a few tenths of a micron here). Rensselaer researchers were actually the first in the world to learn how to evaporate Teflon and are currently exploring adhesion techniques along with additional insulator options, including other polymers and aerogels/xerogels. And the world’s high-tech companies are watching extremely closely.

The “Center” Of It All
As a testament to its pivotal role in the semiconductor industry, Rensselaer is home to the Center for Advanced Interconnect Science and Technology (CAIST). This high-tech research center was established in 1996 with a multimillion-dollar grant from the Semiconductor Research Corporation (SRC), the research arm of the U.S. semiconductor industry, and from New York state, to look at problems the industry will face in the next few years. Rensselaer is also a key member of a major new multi-university Focus Research Center (FRC) in interconnect technology established by the industry in 1998 with significant amounts of matching funds from New York state to host Focus Center New York at the University at Albany. The FRC is concentrated on interconnect problems expected to arise eight to 12 years from now.

Why Are We Obsessed With Interconnects?
Because interconnects play a pivotal role in the semiconductor industry, which is at the heart of the nation’s economic strength. Because new interconnects will transform the computer chips, which are the new backbone of every machine, every business and every industry around the globe, and which have infiltrated all walks of human life. Because we’ve only scratched the surface of what computers can do. As they get faster and more powerful, they make progress possible in everything from communications networks to surgery to environmental remediation.

Why not change the world?