How a spin current flips a memory bit
In a common spintronic memory design, a bit sits in a tiny stack of magnetic materials. One layer is fixed as a reference; another can point in either of two directions. A thin insulating barrier separates them, forming a magnetic tunnel junction—a structure whose electrical resistance depends on how the two magnetic layers are aligned.
To write a bit, the device sends a current whose electrons have a preferred spin orientation. As that spin angular momentum enters the switchable layer, it exerts a small torque on the layer’s magnetization. If the pulse is strong and long enough, the magnetization turns into the other stable direction. The current has changed the bit by transferring angular momentum, not merely by heating or moving charge through a storage capacitor.
Reading without moving the bit
The same junction provides a way to check the result. When the two magnetic layers point in the same direction, electrons can tunnel through the barrier more readily than when the layers point opposite ways. The resulting resistance difference lets the circuit distinguish the two states with a small sensing current.
Because the magnetic direction remains after power is removed, this kind of memory can retain data without continuous electrical power. The engineering challenge is balancing the write pulse: it must reliably switch the magnetic layer, yet stay small enough to limit energy use and avoid disturbing nearby bits. That tradeoff is why the path from microscopic compass to practical memory depends on carefully designed layers, barriers, and currents—not just on making a magnet smaller.
