The science already spawned breakthroughs in computer memory. Parkin’s work in 1989 led to the development of giant magneto-resistive heads that powered a generation of hard disk drives.
Since then Parkin’s work spawned interest in magneto-resistive RAM that some see as a possible replacement for both DRAM and flash. IBM developed an MRAM prototype in 1999, is currently developing a 16 Mbit chip, and has a development effort with Micron on the topic, though Parkin declined to discuss its status.
Separately, Parkin leads work on so-called racetrack memory, another candidate for a future universal memory, one that promises densities that could rival hard disk drives. The racetrack work is going “extremely well,” said Parkin.
“We have some new materials to create more narrow domain walls and thinner racetracks with lower current densities--I call it Racetrack 2.0,” Parkin said, noting it could still take three to five years to prove the technology is feasible for commercial use.
“We need to build prototype devices that operate reliably over a long time and make the characteristics similar so all the racetracks behave the same way,” Parkin added. “Currently we are working on lab prototypes of individual racetracks exploring their physics and materials."
Kevin Roche, lead automation engineer at the IBM lab, keeps the systems running and hosted out tour. The console shows the four vacuum chambers in the $3 million lab.
Click to the next page to see the rest of the slideshow: