DURHAM — For the past 30 years, the Semiconductor Research Corp., or SRC, has quietly funded research crucial to developing next-generation semiconductor chips.
Founded in 1982 by the Semiconductor Industry Association to help counter stiff competition from Japan, to date SRC has funded $1.9 billion in research conducted by a host of universities. In recent years SRC has funded university research at a nearly $100 million-a-year clip, thanks to money provided by its 22 member companies including behemoths such as IBM, Intel and Texas Instruments as well as funding from government agencies.
SRC touts that scientific publications resulting from research it has funded account for 20 percent of the worlds semiconductor research.
That research is overseen by SRCs full-time staff of 42, who have been based in the Triangle since the organizations inception. Today SRC is headquartered off Emperor Boulevard near Page Road in Durham.
And that staff has been headed for 30 years and counting by Larry Sumney, the only president and CEO the organization has ever known. Sumney, 72, recently talked with staff writer David Ranii about SRCs past, present and future. Below are some highlights:
On SRCs formation in 1982: At that time, there was no single company, including IBM, that could (afford) the research necessary to take the next step. So they came up with this idea of a consortium: Lets share the cost. ... Leverage is the key thing. At the same time, university scientists werent focused on semiconductor manufacturing or design issues. Universities did not look at manufacturing as worthy of academic pursuit.
In 1981, university scientists published just 180 scientific papers on semiconductors, versus 2,430 published in 2009. Almost all of those have been SRC (funded) publications.
On locating here: The Triangle beat out Boulder, Colo., Minneapolis, Phoenix and the Silicon Valley as the headquarters site. Robert Noyce, co-inventor of the integrated circuit and co-founder of both Fairchild Semiconductor and Intel, was an advocate for the Triangle.
When it came down to the end, Bob said, you know, what I see is a governor, Gov. Hunt. He injected money into this industry by forming the Microelectronics Center of North Carolina. I see three leading universities: North Carolina State, Duke, UNC-Chapel Hill. We want to fund universities. Thats where you should be. Thats, essentially, the rationale. The Microelectronics Center has since morphed into the nonprofit group MCNC.
On industry advances aided by SRC research: Major themes of SRC-funded research over the past 30 years include reducing the size and cost of semiconductor chips and upping their reliability.
A cell phone back in the 70s, if you could afford one they cost thousands of dollars the receiver unit took up the entire trunk of your car. Now the whole thing is ... something you can stick in your pocket. ... When we started ... the best available storage technology came out of the IBM 3350. In order to do what an iPod does, you would need 126 of these ... just for the memory. In 1976 dollars, that would cost $9 million.
On SRCs decision to fund research in the energy space through its Energy Research Initiative, which it established in 2010: All the technology used to produce an integrated circuit is used to produce a solar cell. So the technology is almost exactly the same. ... The second major application is in the smart grid, where theres all kinds of control mechanisms to accept power coming from your house. If you have a solar panel and youre connected to the grid, you get credit for what you put into the grid. Integrated circuits control all of that such that the grid knows where it came from...how much credit you get.
On SRCs funding philosophy: Our key to being successful, I think, is trying to be as agile as we can to respond to industry changes. And what enables that is we only fund research. We dont fund bricks and mortar at universities, nor do we fund equipment. They can get that stuff from the (National Science Foundation) and from grants from alumni, stuff like that. So we can move almost at the drop of the hat from one area to another area just by changing the contracts with the universities we fund.
On Moores Law, which famously and presciently predicted decades ago that the number of transistors the industry would be able to place on a computer chip would double every couple of years: We are running up against the limits of transistor shrinkage. ... So what is industry looking at? Theyre looking at what were calling More than Moore. Its more than just shrinkage. Its other things. One of the things would be 3-D integrated circuits, where you stack the chips and you connect them.
You might say, well geez, that makes them an awful lot bigger. Not that much. The wafers are so thin that it doesnt enlarge the volume very much.