Biocomputing and neuromorphic computing research may hold the key to improving computer energy efficiency. By taking inspiration from nature’s own efficient systems, such as the human brain, we may be able to address the energy needs of an increasingly digital world.
As computers consume more and more power, scientists are turning to an unlikely source of inspiration for greater sustainability: the humble biological cell. This approach, known as biocomputing, can reduce energy consumption during computing.
A recent article in The Conversation highlights this concept, which leverages nature's own efficient systems to solve one of the most pressing challenges in modern computing. With data centers and household equipment gobbling up around 3% of global electricity demand, and artificial intelligence set to push that number even higher, the need for energy-efficient alternatives has never been greater.
The concept of biocomputing originated from a principle proposed by IBM scientist Rolf Landauer in 1961. The Landauer limit states that the minimum energy consumption required for a single computational task (such as setting a bit to 0 or 1) is approximately 10-²¹ Joules (J). While this number may seem negligible, it becomes significant when you consider the billions of operations performed by computers.
In theory, running a computer at the Landauer limit would make computing power consumption and thermal management irrelevant. However, there's a big problem: to achieve this level of efficiency, the operations must be infinitely slow. In fact, faster computing speeds will inevitably lead to increased energy consumption.
Current processors run at clock speeds of billions of cycles per second and consume approximately 10-¹¹J per bit - approximately 10 billion times the Landauer limit. This high speed is the result of computers working serially, performing one operation at a time.
To solve this energy conundrum, researchers are exploring a fundamentally different computer design based on massively parallel processing. This approach suggests using billions of slower "turtle" processors, each taking just a second to complete a task, rather than relying on a single high-speed "rabbit" processor. In theory, this could allow computers to operate close to the Landauer limit, with energy consumption orders of magnitude lower than current systems.
Web-based biocomputing is a promising implementation of this idea, harnessing the power of biological motor proteins - nature's own nanoscale machines. The system involves encoding computational tasks into nanofabricated labyrinth channels, typically made from polymer patterns deposited on silicon wafers. Biological filaments driven by motor proteins explore all possible paths in the maze simultaneously.
Each biological filament, only a few nanometers in diameter and about a micron in length, encodes information through its spatial position in the maze, becoming an independent "computer." This structure is particularly suitable for solving combinatorial problems that place high demands on the computing power of serial computers.
Experiments have shown that this biological computer requires 1,000 to 10,000 times less energy per calculation than an electronic processor. This efficiency stems from the evolved properties of biological motor proteins, which use only the energy required to perform tasks at the required speed - typically a few hundred steps per second, a million times slower than a transistor.
Recently, significant progress has been made in this area. Heiner Linke, professor of nanophysics at Lund University and author of the Dialogue article, also co-authored a 2023 paper demonstrating the possibility of running computers near the Landauer limit. This breakthrough brings us closer to realizing the potential of ultra-low-energy computing.
While the concept of biocomputing is promising, challenges remain in scaling up these systems to compete with electronic computers in terms of speed and computing power. Researchers must overcome various obstacles, such as precisely controlling biofilaments, reducing error rates, and integrating these systems with current technology.
If these obstacles can be overcome, the resulting processors could solve certain types of challenging computing problems at dramatically reduced energy costs. This breakthrough could have profound implications for the future of computing and its impact on the environment.
As another approach, researchers are also exploring neuromorphic computing, which attempts to simulate the highly interconnected architecture of the human brain. While the basic physical elements of the brain may not be inherently more energy-efficient than transistors, their unique structure and operation offer fascinating possibilities for energy-efficient computing.