ganization, programming, communication, and timing levels. The authors consider neuromorphic computing thoughts around the full compute stack levels of materials, devices, circuits, microarchitecture, system architecture, algorithms, and applications, and propose much tighter interdependence of these levels for future designs. The full compute stack (or design stack) levels in the article are a valuable way to organize neuromorphic computing layers and provide scaffolding to analyze the design and architectural concepts (Schuman 2022). 5. Human brain in silicon Plebe and Grasso published their assessment of efforts to design a computer hardware-inspired brain or a brain in silicon. The authors are not convinced that computing based on these principles will replace conventional Von Neumann approaches. They mention efforts to reverse engineer the brain. In the conclusion section, they state, “Until a theoretical framework emerges to capture essential aspects of neural plasticity and an appropriate technology able to mimic it is devised, the quest for the ‘brain in silicon’ could be severely impaired.” (Plebe 2015) 6. Big data applications of neuromorphic computing Shrestha et al. published their findings on the increase in big data applications and how the capabilities of neuromorphic computing can help meet these needs. They explored neuromorphic computing models and provided an accessible summary of important approaches and comparative details about various developed systems. They compared the TrueNorth, SpiNNaker, Loihi, BrainScaleS, Braindrop, Dynap-SEL, and Tianjic large-scale neuromorphic implemented systems. The neuromorphic computing design choices they benchmarked among these systems were (1) neuron model (Classic leaky integrate-and-fire (LIF), CUBA LIF, Exponential integrate-and-fire (IF)), (2) synapse model (number of weights), (3) implementation choice (digital, analog and mixed-signal, and digital with multiprocessor system on a chip, SoC), (4) architecture (interconnect crossbar size, number of processor cores, memory), and (5) software supported (MATLAB, object-oriented code, Python, PyNN, etc.). Although large differences exist in the systems’ implementation, they all utilize a Von Neumann computing construct. Trying to quantize a fixed number of neurons into a chip is not how a biological brain operates. Brain interconnects take place in a three-dimensional space. Electronics cannot do this. Crossbar interconnects are inefficient and do match the dynamic, programmable, and low-power manner synapses connect. Neuron computational engines are embedded in a fundamentally different as compared to electronics. Simple, streamlined, and optimized neuron computational engines are very different from the CPUs found in electronic computing platforms. (Shrestha 2022). THE AUTHOR James D. Johansen has an interdisciplinary Ph.D. from Liberty University, Lynchburg, VA, USA, in 2019, two master’s degrees in science and religion, and Christian apologetics, from Biola University, La Mirada, CA, USA, in 2015 and 2012, and electrical engineering master’s and bachelor’s degrees from the University of Southern California, Los Angeles, CA, USA, in 1985 and 1982. He is an adjunct professor at Liberty University in their engineering and computational science department, an adjunct professor at the Master’s University, Placentia, CA, USA, a researcher in theoretical biology, and an adjunct professor at Biola University, La Mirada, CA, USA, in their chemistry, physics, and engineering department, a part-time assistant professor at Azusa Pacific University in their engineering and computer science department, and an adjunct professor at Regent University in their graduate school. He has over two decades of experience in systems engineering at two federally funded research and development companies supporting the aerospace industry. He has over 20 conference papers and journal articles, plus a book chapter. He has been a member of INCOSE, IEEE, MORS, ETS, CBS, and EPS professional societies. JOHANSEN Human brain function and the creation model 2023 ICC 315
RkJQdWJsaXNoZXIy MTM4ODY=