Biocomputing questions

A computing system in which biological material performs a meaningful part of storing, transforming or processing information. This site focuses on living-neural systems.

Yes. Peer-reviewed research systems and commercial or remotely accessible platforms exist, although their capabilities remain experimental and specialised.

Cortical Labs markets the CL1 device and provides Cortical Cloud access. FinalSpark provides remote access to maintained human neural organoids for research.

Current pricing depends on the platform and can change. Biocomputers keeps dated price records separate from current official availability so historical launch prices are not presented as live prices.

Yes. Cortical Cloud provides remote access to CL1 hardware, and FinalSpark provides remote access to maintained neural organoids through its Neuroplatform.

Current organisations include Cortical Labs, FinalSpark, university research groups and newer companies developing living-neural systems, interfaces and software.

Some do. Human neural cells in current biocomputers are commonly produced from stem-cell-derived lines and grown as planar cultures or organoids. Other systems use animal neurons or non-neural biological substrates.

Brain organoids reproduce selected features of developing neural tissue, but they lack the anatomy, sensory systems and organisation of a whole human brain.

Some peer-reviewed systems show task-dependent adaptation under feedback. The word learning should be tied to an operational measure such as improved task performance or a persistent change in response.

Published systems have demonstrated closed-loop control, reservoir computing, classification, temporal-pattern generation and experimental responses to drugs. These remain specialised research tasks.

DishBrain reported task-relevant adaptation in a peer-reviewed closed-loop Pong experiment using cultured cortical neurons and electrical feedback.

Cortical Labs has published a CL1 Doom demonstration. The public evidence currently sits at company-demonstration level; the source record contains no peer-reviewed Doom study.

Brainoware is a peer-reviewed brain-organoid reservoir-computing system reported in Nature Electronics in 2023.

Culture lifetime varies with the platform, cell preparation and maintenance system. Current platforms report operation over weeks or months; there is no single lifetime figure for biocomputers as a category.

There is no accepted evidence establishing consciousness in current organoid-computing systems. Neural activity, plasticity and task responses do not by themselves answer the consciousness question.

The biological tissue can operate at low power. A general complete-system energy advantage has not yet been demonstrated on matched workloads once life support and electronics are included.

A research programme that develops and studies computation and learning in brain organoids using neural interfaces, software and closed-loop experimental methods.

A broader research area concerned with engineered living neural systems that sense, process information, adapt and interact with artificial environments.

An informal term for computing systems that use living biological material as part of the information-processing substrate.

A multielectrode array, or MEA, is a set of electrodes used to record electrical activity from biological cells and often to deliver stimulation.

Biocomputing uses biological material. Neuromorphic computing implements neural principles in electronic hardware.

Reliable bidirectional access to large, variable living networks is a central bottleneck, alongside culture lifetime, reproducibility, retention and system-level benchmarking.

Yes, as a research prototype. In August 2026 NUS Medicine, DayOne and Cortical Labs announced a 20-unit CL1 deployment in Singapore. The deployment establishes rack-scale integration; matched performance against conventional compute has not been published.

Yes under suitable research culture conditions. A 2026 Nature study maintained human cortical organoids for more than five years. That result concerns developmental organoids. Five-year biocomputer operation and learned-state retention were outside the study.