|
|
|
Sally McKee
"Most people know that processors are getting faster at a phenomenal rate," says McKee, whose "Hitting the Memory Wall: Implications of the Obvious" was published in Computer Architecture News in 1995. "But memory systems aren’t keeping up. We’ve got a real problem here, and we’re going to be in trouble if we don’t do something about it soon. By the year 2010, it’s not going to matter how fast your processor is—you’ll have to wait for memory to catch up. And to some extent, we’re already there." In the seven years since coining the phrase, McKee has been working in two different directions to improve memory systems: finding high-performance solutions for existing technology and designing new controllers to intelligently optimize a computer’s available memory. In the future, she hopes to expand the medical applications of her work, designing processors that will allow doctors and nurses to spend more time with their patients. To McKee, who has worked for Bell Laboratories, Intel, Microsoft, and Digital Equipment Corporation, it feels like solving a puzzle, and like her hobbies—baking cakes and cooking dinners—it gives her the immediate satisfaction of making people happy. "I’m an experimental scientist," says McKee, who spends her off-hours swing dancing, reading, and going for runs with her dog Stella Bella Hyena. "I like to make things that people like, either to eat or to use or to look at. It’s fun to build things, it’s fun to solve puzzles, and it’s nice when people like what you’ve done. So if I invite you over for dinner, you’re agreeing to be a laboratory animal. But don’t worry—I’ve never had any complaints." |