The effort to make ever-faster computer processors may be facing a looming problem 鈥� but there is a solution.
Historically, we have achieved faster processing speeds by cramming more transistors onto silicon chips. But silicon is limited by Moore鈥檚 Law, which says that the number of transistors ought to roughly double every two years. We have replaced silicon with magnetic solid state hard drives, but they, too, have their physical limitations.
Now, a team of physicists has created a new material and technique to write information 1000 times faster than today鈥檚 best hard drives or random access memories.
Advertisement
A computer hard drive is essentially a series of tiny magnets pieced together. When you save a file, pulses of electric current run through the drive to flip the orientations of those magnets up or down, representing the 1s and 0s of binary code. But these magnets have a cap on how quickly they can reorient themselves 鈥� about a billion times a second 鈥� which limits how fast the drive can save data.
To increase this speed, a team led by at the Czech Academy of Sciences has created a memory storage material out of copper manganese arsenide crystals. These crystals belong to a class of materials called antiferromagnets, in which the spins of adjacent atoms point in opposite directions. This is in contrast to the magnets that make up current hard drives, called ferromagnets, in which all atomic spins are pointing in the same direction.
Outpacing today鈥檚 hardware
The researchers made antiferromagnet squares about the width of spider silk to store each bit of information. They ran a series of extremely short electric pulses 鈥� about a trillionth of a second 鈥� through the antiferromagnet to flip the spins in each square.
They haven鈥檛 yet fired these pulses in quick succession, but if they did, they say they could get the squares to flip about a trillion spins per second, achieving a terahertz writing speed.
This speed isn鈥檛 feasible for ferromagnets. 鈥淚n conventional ferromagnets, it takes increasingly large energies to be able to flip the spins fast,鈥� says at the University of Notre Dame in Indiana. It would take 1000 times as much energy for a ferromagnet to flip at this speed.
Furdyna says this new material shows promise, especially because the experiment was carried out at room temperature. 鈥淢any candidate materials with fancy characteristics exist, but they all occur at low temperatures,鈥� he says.
But it is too soon to say whether this material will be a commercially viable replacement for ferromagnetic hard drives. Antiferromagnets haven鈥檛 yet been optimised for cost or compatibility with existing technology, says Furdyna. Like silicon, which took decades of engineering to reach its potential in smartphones and laptops, antiferromagnets still need years of development, he says.
Reference: 听
