The device would be built with anti-ferromagnetic (AFM) materials.

Researchers at Northwestern University and the University of Messina in Italy have developed a new magnetic memory device. Built of antiferromagnetic (AFM) materials, it solves the limitations of existing hardware for data-centric applications, especially in artificial intelligence. (EXPRESS COMPUTER)

Current day computers and memory chips are unable to keep pace with the increasing demands from big data for bigger, faster, and less power-intensive processors.

Typical, conventional storage methods can get slower with data influx or dissipate too much power in the process – leading to what researchers call a “memory bottleneck.”

The new AFM magnetic memory device from Northwestern and Messina Universities is the smallest ever in size. It also uses exceptionally low electrical current during data-processing. The research on this new memory device will publish in the Nature Electronics journal on February 10.

“Existing hardware cannot sustain the rapid growth of data-centric computing,” said Northwestern’s Pedram Khalili, who led the research. “Our technology potentially could solve this challenge.”

AFM could deliver the best of three worlds

Hugely data-centric applications, such as in AI and machine learning, require enormous computing memory. Ideally, it should be fast, such as static random access memory or SRAM. Further, it should store data like dynamic random access memory (DRAM or Flash). Lastly, it should also expend low energy.

However, AFM materials are fast, secure, and use less energy. They do not need a continuous power supply to retain stored data because they store data using magnetically ordered “spins.” Spins are a kind of quantum mechanical property.

AFM-based devices are also very secure because of their densely packed configuration – which renders them safe from external magnetic fields.

Their density means the devices can be very much smaller. Further, by innovatively using pillars of AFM platinum manganese, the researchers could drastically reduce the size of these devices. With a diameter of only 800 nanometers, they’re about ten times smaller than previous AFM memory devices.

AFM – the icing on the cake

These devices also have another crucial advantage – semiconductor companies can run with their new technology without the need for expensive investment in new equipment.

“This brings AFM memory — and thus highly scaled and high-performance magnetic random-access memory (MRAM) — much closer to practical applications,” Khalili said. “This is a big deal for industry as there is a strong demand today for technologies and materials to extend the scaling and performance of MRAM and increase the return on the huge investment that industry has already made in this technology to bring it to manufacturing.”

Related Story:  Cerebras Systems Launches AI Chip with a Record 1.2 trillion Transistors