MIT uses nanotech to shrink chips to 25nm
Researchers find a way to build cheaper, more powerful processors
Computerworld - MIT researchers have discovered how to use nanotechnology to significantly shrink computer chips, making them cheaper and more powerful.
The new technique will enable manufacturers to produce 25-nanometer chips, which is a huge leap considering that late last year, Intel Corp. made its first move from a 65nm process to 45nm. Intel has already said it hopes to move to 32nm processors sometime in 2009; a 25nm chip would be one level below that.
"The goal is to make smaller patterns on the chips," said Ralf Heilmann, a research scientist at MIT. "The integrated circuits and transistors that are the core of the chips -- the smaller you can make them, the faster they are, and the more you can pack on a single chip. They're cheaper and more powerful at the same time."
Heilmann said that MIT has patented the technique and that different semiconductor companies, which he would not name, have shown interest in it.
The new technique involves nanoscale lithographic technology that makes finer patterns of the lines used to make up the circuits on the chips. Optical lithography uses light to transfer the pattern onto the chip. In the past, Heilmann said, chip makers tried using shorter wavelengths of light to make the lines closer to together. However, that technique has its limits.
Heilmann said the MIT researchers used longer wavelengths to make 200nm lines with 200nm of space in between them. They would then take the same line template and simply move it over, laying down new lines and cut the space between them in half. The researchers did this over and over until they got the lines and spaces down to 25nm.
That gave them four times the pattern density and enabled them to pack in four times as many features such as wires, conductors and transistors.
"If you can speed the chip up while shrinking it at the same time, you get a big performance bump," said Dan Olds, principal analyst at Gabriel Consulting Group Inc. "This looks like it makes shrinking below 32nm work reasonably well from a cost and yield perspective ... and that's a very interesting prospect that has broad implications. The key point is that their process works at the cutting edge -- less than 45nm -- of performance, rather than just being a cheaper way to produce old chips."
Heilmann said he believes the process could be adjusted to produce chips smaller than 25nm. "It's certainly a possibility," he said.
The project team worked in the Space Nanotechnology Laboratory of the MIT Kavli Institute for Astrophysics and Space Research. It received financial support from NASA and the National Science Foundation.
Read more about Hardware in Computerworld's Hardware Topic Center.
- Gartner Magic Quadrant for Client Management Tools The client management tool market is maturing and evolving to adapt to consumerization, desktop virtualization, and an ongoing need to improve efficiency.
- Audit Ready and Asset Optimized: The Solid Promise of an Intelligent Software Asset Management Solution In this paper Frost & Sullivan examines the benefits of enterprise-grade Software Asset Management solutions, and how these solutions serve as the convergence...
- Pragmatic Endpoint Management: Empowering an SMB Workforce in the Age of Mobility Lacking the time for proper training and education, SMB administrators often resort to taking shortcuts to keep their environment running.This paper discusses the...
- Gartner Magic Quadrant for Application Security The market for application security testing is changing rapidly. Technology trends, such as mobile applications, advanced Web applications and dynamic languages, are forcing...
- Redefine Your IT Operations: Remote Office IT Has Never Been Simpler Join us to see why PC Pro named Dell PowerEdge VRTX the "2013 Server of the Year." PowerEdge VRTX may be just what...
- LIVE EVENT: 5/7, The End of Data Protection As We Know It. Introducing a Next Generation Data Protection Architecture. Traditional backup is going away, but where does this leave end-users? All Hardware White Papers | Webcasts