A class of non-volatile memory devices called MRAM, based on quantum magnetic materials, can offer a 1,000x performance increase over current state-of-the-art memory devices. Materials known as antiferromagnets had previously been demonstrated to store stable memory states, but they were difficult to read. This new research also has the potential to establish an efficient way to read memory states, and do so incredibly quickly.
You can blink about 4 times per second. The frequency of this blink is said to be 4 hertz (cycles per second). Imagine trying to blink your eyes a billion times per second, or 1 gigahertz. It would be physically impossible for humans. However, this is the current order of magnitude that modern high-end digital devices such as magnetic memories switch states when performing operations. And many people want to push the boundaries a thousand times more, a trillion times per second, into the terahertz realm.
A barrier to achieving faster memory devices may be the materials used. Today’s high-speed MRAM chips, which are not as common as those found in home computers, use typical magnetic (ferromagnetic) materials. These are read using a technique called tunneling magnetoresistance. This requires parallel alignment of the magnetic components of the ferromagnetic material. However, this arrangement creates strong magnetic fields that limit the speed at which the memory can be read and written.
“An experimental breakthrough over this limit is thanks to another class of materials, antiferromagnets,” said Satoru Nakatsuji, professor of physics at the University of Tokyo. “Antiferromagnets differ from common magnets in many ways, notably because they can be arranged in ways other than parallel lines, which means that the magnetic field resulting from a parallel arrangement can be canceled.” It is believed that magnetization of the ferromagnet is required for the tunneling magnetoresistance to read from the memory, but surprisingly, a special class of antiferromagnets without magnetization is also possible, We expect it to run very fast.”
Nakatsuji and his team believe that switching speeds in the terahertz range are achievable, even at room temperature, but previous attempts required much lower temperatures, suggesting such promising results. However, to improve on that idea, the team would need to improve their devices, and it’s important to improve how they’re manufactured.
“While the atomic building blocks of our materials (manganese, magnesium, tin, oxygen, etc.) are fairly well-known, the way they are combined to form usable memory components is both novel and familiar. No,” said researcher Xianzhe Chen. “We grow crystals into incredibly thin layers in vacuum using two processes called molecular beam epitaxy and magnetron sputtering. is a very difficult procedure, and if we improve it, it will make our lives easier, and we will also be able to produce more effective devices.”
These antiferromagnetic memory devices take advantage of a quantum phenomenon known as entanglement. But nonetheless, the research is not directly related to the increasingly famous field of quantum computing. But researchers suggest that such developments may help or even be essential in building a bridge between the current paradigm of electronic computing and the emerging field of quantum computing.
Original: Approaching the terahertz region
Than: University of Tokyo
