Recent headlines trumpeted the end of Moore’s Law, a 50-year-old prediction that transistors per square inch on integrated circuits would double every two years into the foreseeable future.
And yes, that pattern of exponentially growing computer power has lagged a bit – noticeably, with Intel’s six-month delay last year in releasing 14-nanometer Broadwell chips. Now, the company says its 10-nanometer chips also will be delayed.
To Vanderbilt University Professor of Electrical Engineering Dan Fleetwood, the bigger story is that Moore’s Law, coined by Intel co-founder Gordon Moore in 1965, has held true this long. Doubling every two-and-a-half years instead of every two is still remarkable, he said.
“We usually don’t see, for 50 years, continuous improvement in the performance of anything,” said Fleetwood, who holds the Olin H. Landreth Chair in Engineering and is chair of the Department of Electrical Engineering and Computer Science.
He teaches a popular first-year engineering science class called Moore’s Law and the Microelectronics Industry, aimed at helping students understand the unusual nature of Moore’s Law and how much is still left to be done in the field of transistor technology. Students use a simulation that tracks microelectronics advances with companies’ stock performance and discuss what jobs are available for a variety of engineering majors.
Headlines with phrases such as “the death” or “the end” of Moore’s Law don’t reflect an adage that actually encompassed change from the start, Fleetwood contends.
“Ten to 15 years into Moore’s Law, the number of transistors on a chip went from doubling every year or year and a half to every two years,” he said. “We all know that a reduction in the pace of introducing new technology generations is inevitable. What’s surprising is that it has taken this long.
”For progress to continue to 14 nanometers, it has required heroic efforts on the part of the industry. Ten nanometers is understood. Seven and below is where the research community is. We’re scratching our heads about lots of possibilities and are not sure how those would work.”
Contact
Heidi Hall, (615) 322-6614
Heidi.Hall@Vanderbilt.edu
On Twitter @VUEngineering