According to the MSN website on the 31st, Swiss biocomputing startup FinalSpark has launched an online platform where users can remotely access 16 brain organs. The company's official website states that this Neuroplatform is the world's first in vitro biological neuron platform that allows online access. The company stated that the neural platform is capable of learning and processing information, and bioprocessors like this consume only one millionth of the power of traditional digital processors. Due to its low energy consumption, it can significantly reduce the impact of computing on the environment. Generally speaking, training a large language model like GPT-3 requires about 10 gigawatt hours of electricity, which is about 6000 times the annual electricity consumption of an ordinary European household. If bioprocessors can be deployed, this energy consumption is expected to be significantly reduced. At present, the operation of neural platforms relies on an architecture that can be classified as "wet parts", that is, a mixture of hardware, software, and biological tissues. The main innovation of this platform lies in the use of four multi electrode arrays (MEAs) to accommodate three-dimensional (3D) cell clusters of living tissue - organoids, that is, brain tissue. Each MEA contains 4 organoids connected through 8 electrodes for stimulation and recording. The data is transmitted bi-directional through a digital to analog converter, with a sampling frequency of 30kHz and an accuracy of 16 bits. These key architectures are maintained by MEA's microfluidic life support system and surveillance cameras. FinalSpark has opened its remote platform to 9 institutions to promote biocomputing and research and development. By collaborating with these institutions, we aim to build the world's first live processor. In addition, 36 universities have shown strong interest in neural platforms. In theory, the neural structures that make up a bioprocessor can have a long lifespan, but researchers at FinalSpark have stated that they are currently only "suitable for a few months of experimentation.". With further improvement of the system, the lifespan of organoids is expected to increase. The energy consumption issue of data centers and artificial intelligence models has always been widely criticized. Compared to that, the human brain has 86 billion neurons and 10 trillion synapses, but in terms of energy consumption, it is only equivalent to a few tens of watts. Therefore, in recent years, scientists in the field of information technology have hoped to draw inspiration from biological neural networks to make information processing more efficient and energy-efficient. Brain like computing is such a technology that seeks to design systems similar to biological neural networks structurally. The above attempt directly incorporates organoids into the computing system, providing new ideas for brain like computing. (Lai Xin She)
Edit：GuoGuo Responsible editor：FangZhiYou
Source：People.cn
Special statement: if the pictures and texts reproduced or quoted on this site infringe your legitimate rights and interests, please contact this site, and this site will correct and delete them in time. For copyright issues and website cooperation, please contact through outlook new era email：lwxsd@liaowanghn.com
Recommended Reading Change it 
Home
