# McGill team builds bacterial biocomputer prototype that solves the Subset Sum Problem

- Published: 2026-11-05 07:07 UTC
- Section: Digital & Computing › Computers
- Tags: bacterial biocomputer, microfluidic network, Subset Sum Problem, NP-complete problems, parallel computing, McGill University, biological CPU, Microsystems & Nanoengineering, Dan Nicolau
- URL: https://realtech.media/news/rt-dba6900e1d9ab0d9f0ee9c76_mcgill_team_builds/

**Bacteria moving through a microfluidic network explored candidate solutions in parallel and multiplied as extra processors, the authors report. It is a proof of concept.**

![McGill team builds bacterial biocomputer prototype that solves the Subset Sum Problem](https://sgzwpqqelzmajkcubgkl.supabase.co/storage/v1/object/public/rt-media/2026/11/e10d9122b8549bf9.jpg)
*NASA Marshall Space Flight Center, Wikimedia Commons, Public domain*

A McGill University-led team has built a proof-of-concept computer in which motile bacteria explore a microfluidic network that encodes an algorithm for the Subset Sum Problem, a classic NP-complete problem. The work was published on 11 August 2026 in Microsystems & Nanoengineering [1]. It is the result of a 20-year collaboration among researchers in Canada, Europe, the United Kingdom and Australia.

## How the microfluidic network computes

The resources needed to solve NP-complete problems grow exponentially while problem size grows polynomially. In sequential electronic computers this resource is runtime, and in massively parallel DNA computing it is space, the authors write [1].

The team translated the problem into a graph of possible choices and recreated it as a network of interconnected channels. Each route through the network represents one candidate solution, and junctions acting as "logic gates" represent decision points. Dividing bacteria moved through the channels and collectively explored many routes at once.

The researchers observed where the bacteria travelled and used automated image analysis to trace successful routes. This let them identify the combinations that satisfied the problem's constraints. To prevent overcrowding, bacterial growth can be controlled by adjusting the temperature of the device.

## Bacterial multiplication as added computing power

According to the authors, the computer performs operations combinatorially. The natural multiplication of bacteria, which act as biological CPUs, continuously amplifies computational power, so it grows to match the problem size [1]. "With bacterial division, we have one million chips working at the same time," said Dan Nicolau, senior author and Professor of Bioengineering. He added that each "processor" may perform fewer operations per second than an electronic chip.

A scaling analysis in the paper identifies the point at which biocomputing with multiplying bacteria is expected to outperform electronic computers [1]. The authors also point to low error operation driven by design and a low energy requirement for computing. They suggest that bacterial biocomputation in microfluidic networks has the potential to scale up [1].

The researchers say the device is not meant to replace electronic or quantum computers. They describe it as a resource-efficient, cost-effective tool for problems such as mapping the structure of proteins, and name encryption and molecular biology as possible applications. The reported result is a demonstration on a single problem, the Subset Sum Problem.

## Sources

1. [Microsystems & Nanoengineering](https://doi.org/10.1038/s41378-026-01341-x)
2. [EurekAlert](https://www.eurekalert.org/news-releases/1150117)

---

Source: RealTech, realtech.media — https://realtech.media/news/rt-dba6900e1d9ab0d9f0ee9c76_mcgill_team_builds/