Increasing AI Energy Efficiency With Compute In Memory

Skyrocketing AI compute workloads and fixed power budgets are forcing chip and system architects to take a much harder look at compute in memory (CIM), which until recently was considered little more than a science project. CIM solves two problems. First, it takes more energy to move data back and forth between memory and processor than to actually process it. And second, there is so much da... » read more

Novel NVM Devices and Applications (UC Berkeley)

A dissertation titled “Novel Non-Volatile Memory Devices and Applications” was submitted by a researcher at University of California Berkeley. Abstract Excerpt "This dissertation focuses on novel non-volatile memory devices and their applications. First, logic MEM switches are demonstrated to be operable as NV memory devices using controlled welding and unwelding of the contacting electro... » read more

ReRAM Seeks To Replace NOR

Resistive RAM is gaining renewed attention as demand for faster and cheaper non-volatile memory alternatives continues to grow, particularly in applications such as automotive. Embedded flash has long left designers wishing for better write speeds and lower energy consumption, but as the leading edge of that technology shrunk to 28nm, another problem arose. Manufacturing flash memory at thos... » read more

ReRAMs Look To Silicon For Silicon Compatibility

For such a critical material, silicon oxide is not especially well understood. The semiconductor industry certainly understands how to grow high quality oxides with high breakdown voltages, but what happens in less ideal situations? What does the introduction of microstructure do? If there are regions that are oxygen-rich or silicon-rich relative to the stoichiometric SiO2 composition, how do t... » read more

Efficient Neuromorphic AI Chip: “NeuroRRAM”

New technical paper titled "A compute-in-memory chip based on resistive random-access memory" was published by a team of international researchers at Stanford, UCSD, University of Pittsburgh, University of Notre Dame and Tsinghua University. The paper's abstract states "by co-optimizing across all hierarchies of the design from algorithms and architecture to circuits and devices, we present ... » read more

3 Emerging Technologies: Memristors, Spintronics & 2D Materials

New technical paper titled "Memristive, Spintronic, and 2D-Materials-Based Devices to Improve and Complement Computing Hardware" from researchers at University College London and University of Cambridge. Abstract "In a data-driven economy, virtually all industries benefit from advances in information technology—powerful computing systems are critically important for rapid technological pr... » read more

MEMprop: Gradient-based Learning To Train Fully Memristive SNNs

New technical paper titled "Gradient-based Neuromorphic Learning on Dynamical RRAM Arrays" from IEEE researchers. Abstract "We present MEMprop, the adoption of gradient-based learning to train fully memristive spiking neural networks (MSNNs). Our approach harnesses intrinsic device dynamics to trigger naturally arising voltage spikes. These spikes emitted by memristive dynamics are anal... » read more

End-to-End System for Object Localization By Coupling pMUTs to a Neuromorphic RRAM-based Computational Map

New research paper titled "Neuromorphic object localization using resistive memories and ultrasonic transducers" from researchers at CEA, LETI, Université Grenoble Alpes and others. Abstract "Real-world sensory-processing applications require compact, low-latency, and low-power computing systems. Enabled by their in-memory event-driven computing abilities, hybrid memristive-Complementary... » read more

Using Dynamic Route Map Technique for Insight Into Memristors

New technical paper titled "Empirical Characterization of ReRAM Devices Using Memory Maps and a Dynamic Route Map," from Balearic Islands University, UC Berkeley, Health Institute of the Balearic Islands, International Hellenic University, Technische Universität Dresden, Universidad de Valladolid, and Aristotle University of Thessaloniki. Abstract: "Memristors were proposed in the early 1... » read more

Analog Edge Inference with ReRAM

Abstract "As the demands of big data applications and deep learning continue to rise, the industry is increasingly looking to artificial intelligence (AI) accelerators. Analog in-memory computing (AiMC) with emerging nonvolatile devices enable good hardware solutions, due to its high energy efficiency in accelerating the multiply-and-accumulation (MAC) operation. Herein, an Applied Materials... » read more

← Older posts