
IBM Researcher Stuart Parkin pioneered the development at the company’s Almaden Research Center in San Jose, California, starting in about 2004. Parkin conceived of a device consisting of a city of skyscrapers—each one only hundreds of atoms wide—of magnetic material, with each floor of each skyscraper containing a single bit of data. The data is shot up and down the skyscrapers—almost like a supersonic elevator—by using special currents of electrons for which the spins of the electrons, a quantum mechanical property, are aligned in the same direction. By passing such “spin polarized” currents through the data, the magnetic data can be moved up and down the skyscrapers which are vertical racetracks. These currents are generated by a transistor connected to the bottom of each skyscraper. “In this way, each transistor can store not just one bit of data, as in all other solid state memory, but rather 100 bits,” Parkin said. “This means that one can have a solid state memory with the same low cost of a disk drive but with a performance 10 million times better!”
A personal storage device using racetrack memory could fit into a lapel pin and record every conversation its wearer has for years before filling up. In enterprises, massive storage could be dispersed, with terabytes of information built into every device, sensor, camera and doorknob.
In December 2010, in an article published in the journal Science, Parkin and his co-workers detailed a discovery that puts racetrack memory a step closer to moving from a lab project to a commercial product. The researchers revealed a previously unknown aspect of the physics inside racetrack memory, measuring the movement and processing of digital data as a magnetic pattern on nanowires 1000 times finer than a human hair.
Racetrack memory relies on the spin of electrons to move data at hundreds of miles per hour to atomically precise positions along the nanowire racetrack. The data is stored as magnetic regions, also called domains, in the racetracks. The new discovery allows the precise control of the placement of these domains. By controlling electrical pulses in the device, the scientists can move the domain walls at speeds of hundreds of miles per hour and then stop them precisely at the position needed, allowing massive amounts of stored information to be accessed in less than a billionth of a second.
Parkin and his group have published a series of papers in the past several years in which they have developed a detailed understanding of the dynamics of moving domain walls in magnetic nanowires by using ultrashort pulses of spin-polarized currents. In 2006, they were the first to show that nanosecond long pulses of electric current can push magnetic regions along a wire at 110 meters per second—one hundred times faster than was previously possible.
The speed of data access is what makes racetrack memory so intriguing. Projections are that racetrack can read or write a bit of information in less than 1 to 10 nanoseconds, depending on the length of the racetrack. A hard drives does the same in 3,000,000 nanoseconds. Flash memory takes much longer to write information to memory.
Racetrack memory is also likely to be inexpensive. Solid-state storage today is typically expensive compared to disk drive storage. Each transistor in a solid-state device can store only one bit of information. The only way to pack more information into a flash drive is to pack together the latest tiniest, costliest transistors.
In 2010, a flash drive cost about US$2 per gigabyte. Disk drives cost about 10 cents per gigabyte. Racetrack memory could store 100 bits or perhaps more in a single transistor, which would drop the cost of solid-state storage to the level of disk drives. Racetrack memory could be a successor to flash in handheld devices, and possibly replace hard drives.
Selected team members who contributed to this Icon of Progress:
- Stuart Parkin IBM Fellow, Manager of the Magnetoelectronics and Spintronics Group at the IBM Almaden Research Center, Inventor of Racetrack Memory
- Luc Thomas Research Staff Member, IBM Almaden Research Center
- Masamitsu Hayashi Post-Doctoral Visiting Scientist, IBM Almaden Research Center
- Rai Moriya Spinaps Postdoc, Magnetoelectronics Group, IBM Almaden Research Center