Advanced Electronic Materials · 2015 · 233 citations · 202 references
EngineeringComputer EngineeringComputer SystemsElectronic MaterialsSemiconductor DevicesSemiconductor MaterialsMemory DevicesMemory HierarchyMemory StorageMemory Subsystem DesignMemory SystemsMemristive MaterialsFabricationMemory Cell DesignMemory CircuitsMemory MaterialsNeuromorphic DevicesNeuromorphic HardwarePerformance DegradationVon Neumann Architecture
Digital memory demand continues to rise, yet shrinking cell size is no longer viable due to economic and performance limits, while computing performance remains improving, prompting a review of current memory chip status and its pros and cons from system, fabrication, and materials perspectives. The paper argues for a fundamentally different approach to memory technology, advocating a shift from von Neumann computing toward neuromorphic computing and material implication. The authors review current memory chip status, describe its limitations, reassess proposed solutions, and discuss a transition to alternative computational paradigms such as neuromorphic computing and material implication. They find that present technologies have significant limitations and that previously suggested solutions require reassessment.
The ever‐increasing demand for higher‐capacity digital memory shows no sign of declining. The conventional strategy for meeting such demand, i.e. shrinking of the memory cell size, will no longer be useful at some point in the future, owing to economic reasons and performance degradation. Nevertheless, performance of computing systems will keep improving for the next generation information technology. This indicates the necessity to consider a fundamentally disparate approach to enhance memory technology. Here, the current status of computer memory chips is reviewed and the pros and cons of the present technology are discussed from computing system, fabrication technology, and materials points of view. Based on this knowledge, the limitations of the present technologies are described, and the possible solutions suggested up to now are reassessed. Finally, a shift in the fundamental computational paradigm from von Neumann computing to other alternatives such as neuromorphic computing and material implication, is commented upon.
202
<i>Colloquium</i>: Topological insulators
M. Zahid Hasan, C. L. Kane · Reviews of Modern Physics · 2010 · 19.9K citations · Full text
Dmitri B. Strukov, Gregory S. Snider, Duncan R. Stewart et al. · Nature · 2008 · 11.5K citations
Engineering, Electrical Engineering, Electronic Materials +5
Masatoshi Imada, A. Fujimori, Yoshinori Tokura · Reviews of Modern Physics · 1998 · 7.5K citations