Neuromorphic Chips in the Modern World - Research Paper Example

Technology Forecast: Neuromorphic Chips. Massachusetts Institute of Technology’s Technology Review Journal listed neuromorphic chips in its list of top ten breakthrough technologies of the current decade (Hoff, 2014). According to the MIT journal, neuromorphic chips will be game changers in the electronic technology field because the traditional chips that are currently in use are fast reaching their performance limits. The traditional chips are forcing manufacturers to include more memory caches and transistors in their architecture designs and this requirement always inhibits the chips capabilities. It is for this reason that smart phone devices are ever power stingy. A Neuromorphic chip refers to minute microprocessors which are configured to replicate the work of the human brain. The neuromorphic chips replicate the functioning of the biological neurons that are found in the human brain. The chips provide an insight on how the brain of a rational human being processes information. If the technology is actualized, it would lead to the development of electronic systems that are extremely fast, utilize minimal power and can easily make sense of sensory input and undertake to deliver in real time the tasks that are user defined (Rothenbuhler, 2013). More significantly, devices that are based on the neuromorphic chip will have the ability to interact with the world in exactly the same manner a human being would. Robotic medical devices will be in a position to track a patient’s response to medication over a certain time period and learn to review the dosage limit and catch other problems just like a doctor would. Just by wearing a smartphone gear, it may work with the brain to establish who one is thinking about and in that instance place a call to that particular person. Alarms in the schedule might go off in the expected time even though such alarms were not preset. . Toby Delbruck and Shih-Chii Liu (2012) observe that conventional technology has always been far outpaced by biological efficiency and it is for this reason that neuromorphic engineering designers have set out to develop electronic systems which have a similar computational style with the human biological system. The authors observe that neuromoprhic engineers are already making notable milestones characterized by the ever increasing number of sensory applications and sensors that have the capacity to communicate through signals that resemble the neural spikes in a human brain. Other pointer towards the expected breakthrough and large scale deployment of the neuromorphic chips include the improved performance of novel sensory processing methods. Modern electronic devices are based on the memory dependent Neumann architecture which normally entails moving data between the memory chips and the central processor. This particular approach is highly commendable in instances where sorting out numerals or executing program codes is mandatory. However, the architecture, unlike the neuromorphic design, cannot make sense of image and sound. Neuromorphic design has been put on the spot for its possible ability to dehumanize the future society (Bray et al, 2013). A dehumanized society refers to a society where most of the tasks that are normally undertaken by humans are instead completed by machines. A dehumanized society like the one which is posed by the neuromorphic technology will give a field day to those who are hell bent to use technology to cause harm. Giving machines full human capabilities can scale the acts of terrorism and other forms of human conflicts to a new level. This is because it will now be easier to deploy such machines to go and undertake acts of terror because after all their arrest or death will not have any impact on the traditional human emotions compared to how a particular situation might have turned out if a real human being was used. The cognitive abilities that are attributed to the neuromorphic chips make the technology very secure. The cognitive attributes will ensure that any device that has been fitted by such chips can pay attention to the user’s way of doing things and learn about all the user’s unique habits over time. This cognitive characteristic of the technology makes it easier to assess, prevent, detect, react and recover if the device fitted by the chip is being used by anyone else besides the device’s owner. For example, if a users’ smartphone has since deciphered that the user wakes up a certain time, then it will soon automatically set the morning alarm on at the time the user wakes up. Any action to the contrary, say someone else sets a different time for the morning alarm, then the device can detect that it has possibly been compromised. If a car that had been fitted with the chip suddenly deviates from the normal road that leads home, then the car can automatically detect that it has been stolen. It can either react by switching off or sending a message to the nearby police station. Although neuromorphic chips have been rued as the next avenue of foolproof security against any forms of cyber-attack, this will only be the case if such technology can fully replicate the human DNA and genes sequence. If harnessed well, neuromorphic chips will lead to the era of real human-like artificial intelligence. Already, Qualcomm Inc has manufactured a prototype robot that can use its eyes and arms too separate and pile together physical objects that share similar attributes. In a demonstration at the company’s headquarters in San Diego, the robot that had been fitted by the neuromorphic chips could do with minimal human guidance in separating action figure toys from a pile of toys. Toby and Liu (2012) infer that future advances in the deployment of the neuromorphic based systems will greatly depend on how modules can be systematically combined to come up with the desired function, the eventual morphing of algorithms into hardware and a deeper understanding of the nervous system especially on probabilistic inference methods. Other notable challenges that are inherent in the future application of the neuromorphic technology include those that are linked with the production of silicon designs that are commercially viable (Wang, 2013). Also, the silicone based circuits cannot perfectly replicate the structurally rich and complex cell’s molecular machinery by consistently varying their operating parameters to the prevailing environmental changes. References Bobby D, Dascalu, Sergiu M, & Harris, Frederick C. (2012). Goal-related navigation of a neuromorphic virtual robot. (BioMed Central Ltd.) BioMed Central Ltd. Jayet Bray, Laurence C, Barker, Emily R, Ferneyhough, Gareth B, Hoang, Roger V, Bryant, Monteiro, H. A. P., Murray, A., Reekie, M., & University of Edinburgh,. (2013).Neuromorphic systems for legged robot control. Neftci, E, Toth, B, Indiveri, G, & Abarbane, H. (2012). Dynamic state and parameter estimation applied to neuromorphic systems. (info:doi/10.1162/NECO_a_00293.) MIT Press. Robert, Hoff. (2014). Neuromorphic Chips. Technology Review Journal. MIT. Rothenbuhler, A. (2013). A memristor-based neuromorphic computing application. Shih-Chii, Liu. & Tobi, Delbruck. (2012). Neuromorphic Sensory Systems. Current Opinion in Neurobiology. ­20: 10-18 Sheik, S, Coath, M, Indiveri, G, Denham, S L, Wennekers, T, & Chicca, E. (2012).Emergent auditory feature tuning in a real-time neuromorphic VLSI system. (info:doi/10.3389/fnins.2012.00017.) Frontiers Research Foundation. Tseng, K.-C., Viterbi School of Engineering, & Parker, Alice C. (2012). Neuromorphic motion sensing circuits in a silicon retina. Los Angeles, California: University of Southern California. Wang, R., Wang, Runchun, University of Western Sydney, & MARCS Institute. (2013).Neuromorphic implementations of polychronous spiking neural networks. Read More