The transistors in your laptop switch on and off about three billion times per second. They do it using electrons, tiny charged particles zipping through silicon at speeds that would make your head spin. Nobody questions this. It’s just how computers work.
Except some engineers decided to ask: what if it didn’t have to be?
Fluidic computers, machines that process information using the movement of liquids rather than electrical current, have existed in various forms since the 1960s. Early versions showed up in industrial control systems, aircraft, and Early versions showed up in industrial control systems, aircraft, and military hardware of the era. They worked.. They worked. They just couldn’t compete with silicon once the transistor revolution took off. So the idea got shelved, and most engineers forgot it ever existed.
Recently, a small group of researchers built something that brought fluidic logic back, not as a novelty, but as a serious engineering project aimed at a specific problem that modern computers handle badly. What they built uses water moving through microscopic channels to perform logical operations. And here’s the part worth sitting with: the environment where it works best is the one where conventional electronics either overheat, break down, or simply refuse to function.
When Electronics Go Where They Shouldn’t
Think about the inside of a human body. A deep-sea drilling rig. The reactor vessel of a nuclear plant. These aren’t edge cases. Engineers desperately need computational control in all three, and the standard answer, just use a chip, creates serious problems in every one of them. Electronics and extreme heat don’t mix well. Electronics and strong magnetic fields don’t mix at all. And electronics inside a living body trigger immune responses, corrosion, and a long list of headaches nobody has fully solved.
Quantum computers, which many people assume will eventually replace silicon for hard problems, actually make the environmental problem worse, not better. Most quantum processors require cooling to temperatures near absolute zero, colder than outer space, to maintain the fragile quantum states they depend on. They are the opposite of a machine you can slide into a warm, wet, electromagnetically noisy environment and expect to function.
A fluidic computer has no electrons to disrupt. No transistors to fry. No quantum states to collapse. It runs on pressure differentials and the physical geometry of tiny channels. A strong magnetic field doesn’t care about it. A body temperature of 98.6 degrees doesn’t bother it. It just keeps moving water through its channels and doing its job.
The Logic of Water
The underlying principle is simpler than it sounds. Binary logic, the foundation of every computer ever built, works by distinguishing between two states: on and off, one and zero, yes and no. In a silicon chip, that distinction is made by whether electrons are flowing. In a fluidic system, it’s determined by whether fluid is flowing and which path it takes.
Tiny valves and channel geometries can force a stream of liquid to behave like a logic gate. Water flowing one way means “true.” Water blocked or diverted means “false.” Chain enough of these decisions together, and you have a machine that can process information. It’s slower than a modern processor by a wide margin. Nobody is building a fluidic gaming computer.
But slow is fine when the alternative is nothing. In environments where conventional chips fail completely, a system that processes at a fraction of the speed is infinitely better than one that can’t operate at all. The applications researchers are most focused on involve biological and chemical sensing, systems that need to sit inside a living body or a corrosive industrial environment for months, reading data, making decisions, and reporting back without a battery, without a wireless signal, and without ever generating heat or electromagnetic interference.
And here’s the strange part: the same properties that make fluidic systems unattractive for general computing, slow, bulky by nanoscale standards, and limited in complexity, turn out to be features in these specific contexts. A system that can’t run hot can’t damage tissue. A system that doesn’t generate electromagnetic fields can sit next to an MRI machine. A system that runs on fluid pressure can, in theory, harvest energy from the biological movement around it.
What Quantum Still Can’t Do
Quantum computing gets most of the attention when people talk about the future of computation. And for certain categories of problems, cryptography, molecular simulation, optimization at massive scale, the theoretical advantages of quantum systems are real and significant.
Here’s the thing. Quantum systems are fragile in a way that rarely gets discussed outside the lab. Any stray vibration, a slight temperature bump, a passing electromagnetic field, any of it can cause the processor to lose coherence. The quantum states collapse into noise. The computation fails. Engineers call this decoherence, and at places like Google’s quantum lab in Santa Barbara, managing it consumes more engineering effort than almost anything else in the field.
Fluidic computing doesn’t compete with quantum. It goes where quantum cannot go and won’t go anytime soon. Into bodies. Into blast furnaces. Into settings where the machine needs to be tough rather than clever, and still running six months later. There’s a version of the future where quantum processors crunch the hard math in a cold, clean data center, while fluidic computers handle the wet, warm, messy work of operating inside the physical world. That’s not a consolation prize. That’s a different job entirely.
That division of labor sounds strange until you think about how biology already solved the same problem. Your brain runs on electrochemical signals, essentially a fluidic logic system, while the muscles it controls operate on mechanical principles entirely. Nature figured out long ago that different environments need different computing substrates.
The engineers working on water-based computers are arriving at the same conclusion from the other direction.
Whether fluidic logic ever moves from laboratory curiosity to widespread deployment depends on problems nobody has fully solved yet, miniaturization, programming complexity, and integration with conventional systems. But the core claim is already demonstrated. Water can compute. It just took silicon dominating everything for sixty years before anyone thought it was worth proving again.
