Brain Cells Could Replace Silicon Chips—Here's Why That Could Save Energy

Biological computers might speed processing tasks

  • Brain cell computers might be more energy efficient than silicon chips. 
  • Biological computers could one day be faster than traditional models. 
  • Researchers have demonstrated that DNA computers can be used for storage.
A brain growing out of a computer chip with wires running from it.

Just_Super / Getty Images

Your computer might one day run on brain cells. 

Researchers recently found that biological computers could save electricity over the silicon chips we use today. In the paper announcing the discovery, scientists described using 50,000 brain cells grown from stem cells. It's part of a growing movement to harness the power of evolution to make computer processing faster. 

"This is fueled by the impressive capabilities of a brain compared to a computer, such as its efficiency, continuous learning, and intuitive decision-making," Thomas Hartung, a professor at Johns Hopkins University and one of the lead authors of the paper, told Lifewire in an email interview. "The fastest supercomputer in the world matched the estimated computational capacity of a single human brain only in June last year, but this computer occupied 6,800 sq ft and cost $600 million."

Making Biological Computers Work

The research team described a method for creating "organoid intelligence," or OI, in their paper. The approach would use brain organoids grown in cell culture that share some of the same functions and structure as brains. 

The idea of using brains as a computer isn't entirely new. Hartung noted that his fellow researchers showed last year that a brain cell culture could learn to play the computer game, Pong. Other scientists have demonstrated how to control robots with brain organoids.

"We are finding out how to optimize such processes and use them as benchmarks for optimizing the models or study substance effects," Hartung said. 

But getting brains to work as computers is tricky. James Giordano, a professor of neurology and biochemistry at Georgetown University Medical Center, said in an email interview that one approach is computer fusion that involves two basic architectures: the first is where computational hardware is structurally "fit into" a neural array of cells. The second method is where nerve cells are fitted onto computational circuits, individually or in small clusters. 

"Such approaches can be used to enable computational systems to 'interpret' the activities of nerve cell nodes and networks, as well as whole systemic organoids; and in the latter case, can enable properties of nerve cells to be accommodated and transmitted directly onto computational systems," he added. 

I can imagine rather that we learn how the brain works from attempting biological computation and model our computer architecture accordingly.

A Means to an End 

Biological computers could eventually take over some traditional computing tasks, Dave Turek, the chief technology officer of CATALOG, a company developing DNA-based data storage and computers, said in an email. He pointed out that in recent years, the IT industry has witnessed a proliferation of purpose-fit technologies, including accelerators like GPUs, quantum computers, and powerful parallel computers. 

"This performance and scale, however, come at the expense of higher energy consumption, larger memory and long-term storage demands, and higher management complexity," Turek said. "This has generated tremendous interest and momentum in chemistry-based DNA computing systems, which have a far smaller physical footprint, consume orders of magnitude lower energy, and are resistant to traditional electronic security vulnerabilities."

Turek's company recently achieved a DNA-based computation milestone by demonstrating fundamental parallel search capability using DNA chemistry. CATALOG encoded about 17,000 words from Shakespeare's Hamlet into DNA in September. 

"CATALOG's innovative approach shows, for the first time, how to leverage the massive parallelism of DNA chemistry to search almost any amount of data stored in DNA without the expected proportional increase in resources," Turek said.

A graphical brain model displayed on a computer screen with a person in the background reclining in a chair.

gordonenkoff / Getty Images

Despite recent advances, don't expect to buy an organic computer for your desktop soon. Hartung said biological computing is more a "vision" than a practical reality. 

"A laptop, which is fed regularly with cell culture media, is difficult to imagine," he added. "Personally, I can imagine rather that we learn how the brain works from attempting biological computation and model our computer architecture accordingly."

Petr Sulc, an Arizona State University researcher in biomimetic nanotechnology and molecular computing, said in an email that he expects there will soon be uses for cell and biomolecule-based computation for diagnostic and therapeutic purposes.

"The use of brain-cell organoids to carry out more energy-efficient computation might still be quite far away," he added. "Besides the appeal of 'energy efficacy,' one would need to take into account the high cost associated with producing and maintaining these cells, which could outweigh the computational efficacy such a system would offer."

Update 3/9/2023: Corrected the source's name in paragraph 3.

Was this page helpful?