The Grill: Caroline Ross

Working at a microscopic level, Caroline Ross actually creates and modifies MRAM, or magnetoresistive random-access memory. When we store our birthday photos and thesis papers on our PCs, our thought process usually doesn’t go further than hitting the Save button. We put a large amount of trust in our computers’ storage systems. Using the science of magnets, Ross, a professor at MIT, is modifying RAM to help increase the performance of our favorite machines.

Dossier

Name: Caroline Ross

Title: Professor

Organization: Department of Materials Science and Engineering, MIT

Location: Cambridge, Mass.

Favorite technology: “My favorite toy has to be my Mac.”

Dream dinner guests: Family and friends.

What makes you laugh? “Monty Python, and NPR’s Wait Wait . . . Don’t Tell Me!”

Favorite book: “I have several favorites; I mostly read history, such as Barbara Tuchman’s books, and stories like the Sherlock Holmes books.”

What is a thin film? A thin film is just a thin layer of something — a metal or a semiconductor or an insulator, for example. I work mostly with thin layers of magnetic metals such as cobalt, or magnetic oxides such as iron oxide — rust. Multilayer magnetic thin films are particularly interesting because their electrical resistance depends on their magnetization state. This phenomenon, called magnetoresistance, allows us to use electrical measurements to detect the magnetic configuration of the thin film and therefore to read back data stored in the film.

How is the science of magnets changing the way computers will store data in the future? Computers already use magnetism to store data; the hard disk in your computer stores data as magnetization patterns in a magnetic thin film. However, hard disks are relatively large, and the moving parts can lead to reliability problems. MRAMs provide an alternative data storage mechanism that does not have any moving parts and that can easily be integrated into computers. There are some new discoveries in magnetism that are very important in MRAM — for example, current-induced magnetization reversal to write the bits, magnetoresistance to read back the bits.

How are nanomagnets involved with MRAM? A nanomagnet is a tiny magnet, with a length less than a micrometer. It turns out that when magnets are shrunk to small dimensions, they behave differently compared to macroscopic magnets. Nanoscale magnets often have very simple magnetization states, which makes them well suited to storing digital data. So in MRAM, each bit of data is stored in one nanomagnet, which is addressed using one transistor.

What’s different about MRAM? MRAM is special because it is nonvolatile, fast and dense. Nonvolatile means it stores the data even when the power is turned off — unlike the DRAM in your computer, which loses its data unless it is constantly refreshed. There are other types of nonvolatile memory — for example, flash memory — but MRAM has advantages over flash in terms of its speed and its endurance.

What does this mean for computers? Some people call MRAM a “universal memory” because it could replace all the memory in your computer — the DRAM, SRAM, flash and hard drive — with one type of memory. Probably the most obvious difference would be that when you start up the computer, you would not need to wait for all the programs and data to be copied from the hard disk onto the RAM, so the computer would switch on instantly.

I understand that MRAM isn’t commercially available yet. Do you foresee it in the mainstream market anytime soon? There are several commercial products. For example, Everspin, a spin-off from Freescale, sells products for industrial automation, storage systems and space systems. However, the market at the moment is very small compared to the market for other memories. It is likely that MRAM will continue to move into niche markets as the manufacturing processes improve. We will not see it in large-scale use until the cost decreases.

Where does your passion for this field of research come from? Did you want to change the world, or did you just kind of fall into it because you were good at math and science? I was always interested in making very tiny things. And magnetism on the small scale is fascinating, so this project represents a good combination.

How did you end up at MIT, and where were you before that? I’m from London. I did my B.A. and Ph.D. in Cambridge, U.K., and came to Harvard as a postdoctoral fellow. Then I worked in a hard disk company in California for six years as an engineer before coming to MIT.

What do you consider your greatest accomplishment thus far? In my work here, it is probably the development of new ways of creating small structures by taking advantage of self-assembly processes. This could allow us to keep reducing the size of integrated circuits and other devices.

Do you have any advice for young women just starting out in the field? There are certainly differences in how men and women are perceived. Science and engineering is stereotypically male, and when I worked in industry, I was the only female Ph.D. in the company when I joined. However, women are slowly making inroads, and I always try to encourage young women to go into science and engineering. It is a lot of fun.

What is enthralling about it? Can you elaborate on why you do this type of work and what motivates you? It is fun to decide what to study and to design an experiment to investigate it. It’s fun to be the first person to know something. It’s great to have a small group of students and postdocs working together so that new ideas come out of their collaborations.

When you’re not changing the world one thin film at a time, what are you doing during your free time? I have a sailboat, which I race, and I do tae kwon do. But most of my time is spent with my 4-year-old daughter. I’ve been sailing for 20 years. As for tae kwon do, I’ve done it for about 20 years too, and yes, I’m a black belt.

— Interview by Sara Forrest, a freelance photographer and writer in New York (studio@saraforrestphoto.com)