IBM and US Air Force collaborate to make a neurosynaptic supercomputer
395
views
The artificial intelligence evolution is being embraced by the US Airforce in collaboration with IBM.
The complexity of the human brain is still not fully understood. Our ability to process multiple layers of evolving information is vast. With 100 billion neurons and 100 trillion synapses, the human brain has a processing capacity that no computer could rival. However, IBM In collaboration with the US Air Force Research Laboratory (AFRL) intends to create ‘the first-of-a-kind brain-inspired supercomputing system.’ The system will be powered by a 64-chip array of the IBM TrueNorth Neurosynaptic System. (IBM, 2017)
IBM is currently building a scalable platform for the AFRL which will feature an ‘end to end software ecosystem designed to enable deep-neural-network learning and information discovery’ (IBM, 2017). The system will have the processing capacity equivalent to 64 million neurons and 16 billion synapses and will consume the same energy as a 10-watt lightbulb (IBM, 2017)
loading...
Artificial Intelligence and deep neural networks have allowed for advancements in voice recognition and ‘real time’ interpretations of human behaviour. In the world of business AI has also been utilised to understand consumer trends and purchasing patterns. Moreover interpreting emotions and giving insights to a plethora of sensory data is central to deep neural networks and AI.
The collaboration with the US Air Force will undoubtedly have connotations for national security – perhaps using deep learning to predict conflict, weather patterns and global warming. Also, such technologies could see more user-friendly, intelligent graphics interfaces, that can read the emotions of the user. AI and deep learning may also see advances in the performance of hardware that can adapt accordingly to environmental factors or anticipate changing circumstances. This is likely to be the beginning of many similar collaborations in the field of artificial intelligence.
Loading...