The semiconductor industry is the only field I know where "business as usual" means manipulating matter at the atomic scale and still getting better performance quarter after quarter. While other industries slap a fresh coat of paint on last year’s product and call it innovation, silicon foundries are out here bending the laws of physics. Quite literally.
We’ve now reached the point where transistors are just a few atoms thick. Not figuratively—actually, atoms thick. And with that comes a new set of headaches that no amount of marketing spin can cover up. This is where traditional physics steps out of the room, and quantum mechanics walks in with a smirk.
When transistor gates shrink to atomic dimensions, quantum tunneling is no longer some obscure effect—it becomes a primary limiting factor. Electrons stop caring about barriers, start doing as they please, and degrade gate performance unpredictably. And if that wasn't enough, gate oxides suffer deterioration at the atomic level, where even a single missing or misplaced atom (a point defect) can throw off performance—or completely kill—a chip design hosting over a billion transistors.
And this brings us to the most interesting frontier in this atomic revolution: 2D transistors.
Graphene, the poster child of 2D materials, flips everything we thought we knew about conductivity. In the 3D world, conductivity is a function of cross-sectional area and length. But graphene doesn't play by those rules. Electrons flow ballistically, with virtually zero scattering. In theory, this means crazy high mobility.
Other 2D materials have teased us with even more tantalizing properties—like superconductivity under the right conditions. Now, you might ask, if a single 2D transistor is so incredible, why aren’t we building full chips with them? Good question.
The reality is, making a single 2D transistor in a lab is relatively straightforward. But scaling that up to thousands—let alone billions—of transistors? That’s where the wheels fall off. Material quality becomes an issue at scale. A single grain boundary, defect, or foreign atom can throw a wrench into the works. Uniform doping, consistent thickness, and the ability to reliably produce both N-type and P-type variants? That’s a chemistry nightmare and a fabrication headache rolled into one.
At this stage, designing complex circuits with 2D transistors is like trying to build a cathedral out of wet tissue paper. It's possible in theory, maybe even provable on paper. But in practice, it's fragile, unreliable, and entirely impractical for commercial-scale production.
Scientists in China have recently developed the most complex microprocessor to date made from a two-dimensional (2D) material. The new device, called RV32-WUJI and made from molybdenum di-sulfide, consists of 5,931 transistors and is 3 atoms thick.
The researchers behind the new device, led by Professor Wenzhong Bao from Fudan University in Shanghai, have been able to integrate the transistors into a functional processor. The processor, which is 6mm x 6mm in size, is capable of executing 32-bit instructions and has a yield of 99.7% when manufactured. To achieve this, the researchers utilised machine learning to optimise each process step in the fabrication of the device.
The processor has been designed with a RISC-V core, which is an open-source architecture that allows for a wide degree of customisation and modification. The use of RISC-V also helps to enable the development of future processors based on the technology.
While the new device is still in its infancy, the researchers believe that it has the potential to be used in edge-computing applications. These devices, which are often found in IoT devices, are responsible for processing data locally, which can help to improve data privacy and security.
The researchers are also planning to explore the use of the new technology in other applications, including smart sensors and chips for the Internet of Things. The use of 2D materials in these applications could help to provide a more cost-effective and energy-efficient alternative to current silicon-based technologies.
Whenever a shiny new microfabrication breakthrough hits the headlines—especially ones involving novel materials or transistor structures—I’ve only got one real question: Can it be used to build an actual CPU? Not a one-off switch or a lab-bound curiosity, but something that runs logic, processes data, even in its most primitive form. That’s the litmus test for me.
And in the case of RV32-WUJI, the researchers at Fudan University have done exactly that. They’ve taken 2D molybdenum disulfide (MoS₂) transistors and successfully assembled them into a working microprocessor. That’s a huge deal. Not because it’ll end up in your smartphone next year—it won’t—but because it proves something we’ve needed to see: that 2D semiconductors aren’t just interesting—they’re viable.
The specs are modest. At just under 6,000 transistors, this isn’t even in the same league as a basic ARM Cortex-M0. It won’t be driving your IoT device or running your smart fridge. And that’s okay. It’s not about performance—it’s about possibility.
What matters is the fundamental capability of MoS₂ transistors. Compared to silicon, they offer a higher bandgap, enabling lower leakage currents. They switch faster. They’re more energy-efficient. These aren't minor advantages—they’re critical traits for the future of logic devices. If this material scales, we could see CPUs and even GPUs that deliver higher performance per watt, extending battery life in mobile devices or slashing power requirements in edge compute applications.
This processor isn’t the answer yet—but it’s one hell of a proof of concept.
In essence, RV32-WUJI isn’t a destination—it’s a signpost. It’s showing us that complex logic made from 2D materials is more than just an academic exercise. It’s real, and it’s coming. Whether it’s microcontrollers, flexible wearables, or ultra-efficient compute at the edge, this class of devices is laying the groundwork.
So, is this the future of 2D microcontrollers? Not quite. But it might just be the spark that lights the way.
