A team at the University of Edinburgh has found a mathematical way to switch magnetic memory bits using up to a hundred times less energy. The method, presented in the journal Advanced Materials, could sharply cut the power draw of future memory chips – for now, however, only in simulations.
Optimal control shapes the switching pulses
The group led by Elton J. G. Santos, together with Mohammad H. Badarneh and PeiYu Cai at the University of Edinburgh’s Institute for Condensed Matter Physics and Complex Systems, applies so-called optimal control theory (OCT) to magnetic memory. Instead of flipping bits with standardized magnetic-field pulses, the framework computes the ideal shape of a pulse that reverses the magnetization deterministically within a few picoseconds. The tests use two-dimensional van der Waals magnets – ultrathin layered crystals seen as candidates for future data storage.
Up to two orders of magnitude less energy
According to the calculations, switching energy drops by up to two orders of magnitude compared with conventional field protocols, while the required field amplitude is more than ten times smaller. The achievable values sit in the femtojoule range, matching or beating current-driven schemes such as STT- and SOT-MRAM. The authors place their concept close to the Landauer limit, the thermodynamic minimum for processing a single bit.
- Switching time in the picosecond range
- Switching energy up to 100x lower than conventional field pulses
- Benchmarked against DRAM, STT-MRAM and SOT-MRAM
Still pure theory – but with a blueprint
An important caveat: the results come from modeling and simulations, not from a fabricated chip. The study does, however, offer concrete implementation guidance, including optimized device designs and methods for delivering the magnetic fields. The approach can also be adapted to electrical currents and ultrafast laser pulses. If the principle holds up experimentally, it would be especially relevant for energy-hungry data centers and memory-heavy AI systems, whose power consumption is currently rising fast.
Sources: ScienceDaily · Advanced Materials (DOI 10.1002/adma.202523059) · University of Edinburgh Research Explorer
