Being explored as a potentially more energy-efficient alternative to silicon-based processors, biocomputing uses biological materials like neurons for carrying out computation.
Some view it as a promising direction for handling the growing energy demands of large-scale AI systems, while others remain skeptical.
Swiss startup FinalSpark is a known name in this emerging field. The company operates what it describes as the world's first remotely accessible platform for biocomputing research.
Let's see what they are up to.
FinalSpark: What Do They Do?
A demo of the FinalSpark Neuroplatform.
FinalSpark grows human neurons from stem cells and connects them to electrode arrays. Researchers can access these biological neural networks remotely to conduct experiments via Neuroplatform.
The company claims its bioprocessors consume significantly less power than traditional processors. Since starting biocomputing work, they have tested approximately 10 million neurons and collected 20 terabytes of experimental data.
The platform supports both electrical and chemical stimulation methods. Users can programmatically trigger neurotransmitters like dopamine, glutamate, and serotonin through a Python-based stimulation API.
Their Neuroplatform offering currently provides services to many universities, which include the likes of the University of Michigan, Free University of Berlin, University of Exeter, Lancaster University Leipzig, University of York, and a few others.
They have a live demo page where visitors can observe real-time electrical activity from living neurons on their biochips. On GitHub, they maintain open source tools and code repositories related to Neuroplatform.
And, for those interested in biocomputing, FinalSpark also runs a Discord community with over 1,000 members.
Suggested Read 📖
Sourav Rudra
