This simulation shows how crystalline nuclei expand in just 600 picoseconds, converting a phase change material from a glasslike state to crystalline. F. Rao et al./Science

Computer manufactures are always looking for the next best technology, and thanks to an impressive bit of materials science, a little-used form of random access memory (RAM) could soon be taking over the insides of your personal computer. That’s because materials scientists in China have recently found a way to speed up—by more than a factor of 10—so-called phase-change random access memory (PCRAM), which can hold onto information even when your computer’s power is off.

For decades, computers have been getting smaller, faster, and cheaper. But advances in memory technologies have slowed in recent years, prompting researchers to consider alternatives. Dreamed up in the 1960s, PCRAM has always been a promising candidate for RAM—the rewritable scratch pad that a computer’s central processor uses while making its calculations. But it was much slower than most common forms of RAM—including static RAM (SRAM) and dynamic RAM (DRAM)—which can hold information only temporarily.

To serve as computer memory, a gizmo must be able to faithfully record either a 0 or a 1. PCRAM does this by switching between two states (hence, the phase-change name): a regimented crystalline order that allows for the easy flow of electricity and a glasslike jumble of atoms that does just the opposite. PCRAM records one number when it’s in a high-conductance crystalline state and another when it’s in a low-conductance glass state. Sending a relatively large current through the material heats the bit and changes its state, writing or rewriting the data.

The problem has been that the most commonly used phase change material—an alloy of germanium, antimony, and tellurium known as GST—can be an uneven performer, sometimes switching from amorphous to crystalline in 10 nanoseconds, which is just as fast as the DRAM in today’s computers. At other times, it can take hundreds of nanoseconds to write data bits, which is too slow to be competitive. Now, however, Feng Rao, a materials scientist at the Chinese Academy of Sciences’s Shanghai Institute of Microsystem and Information Technology, and colleagues have found a way to write all PCRAM bits quickly, making it faster than most alternatives, including NAND flash—one of the other kinds of memory that’s able to store information absent a power supply.

Nearly a decade ago, researchers working on GST realized that even in its amorphous state, the phase-change material could harbor crystalline precursors, cubic arrangements of atoms that can act as nuclei for switching to a full-fledged crystal. So, Rao’s team carried out systematic computer calculations to see whether adding different metal atoms to the GST mix would help stabilize nuclei at the elevated temperatures that take place during the phase switching.

Those simulations, Rao says, revealed that “scandium is the key.” Adding the element creates strong bonds with neighboring antimony and tellurium atoms, forming cube-shaped nuclei that remain intact even when enough electricity is zapped through the material to raise its temperature to 600 K, which promotes a fast switch between the amorphous and crystalline phases. And when the researchers synthesized their new phase change material, they found that the nuclei consistently caused the material to switch between the two states in less than 1 nanosecond, they report this week in Science.

Hongsik Jeong, a PCRAM expert at Tsinghua University in Beijing, says the new speed boost is “almost the ideal value for storage class media applications.” In fact, even without the speed boost, PCRAM has just begun to be used in some data center servers. So, the souped-up version could soon expand its reach. However, Jeong adds, the faster PCRAM must still prove that it can be scaled up, withstand the high temperatures found in standard chip-manufacturing conditions, and still be able to rewrite bits of data many trillions of times to match DRAM’s performance. If PCRAM can pass those hurdles, the trend toward smaller, faster, cheaper computers will likely continue well into the future. (Science)