SOURCE-LINKED INTELLIGENCE

Symmetry-driven Writing with Intrinsic Torques for Computing beyond Hardware limits

CORDIS · observation · Publication date unknown

Symmetry-driven Writing with Intrinsic Torques for Computing beyond Hardware limits "The explosive growth of artificial intelligence has exposed a fundamental flaw in conventional computing: the von Neumann architecture is inefficient for data-intensive brain-like tasks. A solution lies in unconventional computing architectures that integrate memory directly with processing. The key enabling component is a new class of ultra-fast, high-density, non-volatile memory. While magnetic random-access memory is a leading candidate, its progress has slowed. The incumbent switching technologies, spin-transfer torque and spin-orbit torque, are too energy-intensive, creating a critical roadblock. A transformative solution lies in a new class of materials: Non-Collinear Antiferromagnets. These materials offer a unique ""best-of-all-worlds"" combination: they retain the robust stability and ultrafast dynamics of antiferromagnets, but also possess produces large electrical signals for readout. With efficient

Read original source ↗ Open in workspace

recordType

award

status

SIGNED

region

EU

value

202125.12

unit

EUR