It’s stunning to think how much computer chip miniaturization has progressed in a few decades. Take the best, most advanced computers from thirty years ago and they look nothing like the device you’re likely reading this on. This progress has been summed by Moore’s Law, which states that the number of transistors in a dense integrated circuit (IC) doubles about every two years.

Of course, Moore’s law isn’t really a law, it’s an estimate — although a remarkably accurate estimate. But the estimate only stands until you get to really, really small sizes. When your transistors become very small and you enter the quantum world, the “normal” laws of physics don’t apply anymore. We’re reaching that stage.

In a study published in Nature, researchers detail the construction of a transistor gate that’s 0.34 nanometers (nm) long — about 4 times the length of a carbon atom.

A transistor is essentially a device that amplifies or switches electrical signals and power. The gate is the component that switches the transistor on and off; think of it as the “yes” or “no” controller of the transistor. Previously, researchers managed to get transistor gates to lengths of 1 nanometer and even below, but this is pretty much as low as it gets.

“In the future, it will be almost impossible for people to make a gate length smaller than 0.34 nm,” the paper’s senior author Tian-Ling Ren told IEEE Spectrum. “This could be the last node for Moore’s Law.”

While previous ultra-small transistors used carbon nanotubes for the gate, Ren and colleagues decided to opt for graphene, which is essentially a sheet of carbon so thin that it behaves like a 2D material. They started out with a layer of silicone dioxide as the base structure, and then used vapors to deposit graphene on top of the silicon dioxide. They then sandwiched the graphene with aluminum oxide, essentially cutting off its electrical properties from the rest of the transistor. Then, they etched a step in the sandwiched materials, exposing the edge of the graphene sheet to the vertical wall of the step, essentially creating an atomically thin transistor gate.

Of course, this is a proof of concept. There’s still a long way to go before we incorporate this type of technology into working microchips, but the fact that scientists have gone to the absolute smallest sizes possible is remarkable. But it also speaks to the physical limitations future microchips will face. What will you do if you can’t fit more transistors onto a board?

Before you freak out, Moore’s law isn’t driven only by transistors, improvements in architecture and software can also drive progress. But researchers looking to make more powerful computers have their work cut out for them.

The study has been published in Nature.