Good for 2D semiconductors. imec puts "atomic channels" on their technology roadmap in 2036. Many puzzles to solve till then! There is growing consensus that atomically thin 2D semiconductors are the main post-Si option. CFET = stacked NMOS and PMOS. tomshardware.com/news/imec-reve…

2024 Big Race is TSMC N2 and Intel 18A semiwiki.com/semiconductor-… #Chiplets #Intel18A #Samsung3nm #TSMCN2

Computing, since its inception, has been processor-centric, with memory separated from compute. Inspired by the organic brain and optimized for inorganic silicon, NorthPole, the new chip from IBM, is a neural inference architecture that blurs this boundary.

A nice collaborative work (KCL, U’Edinburgh & U’Southampton) led by Mari Napari. An ultrathin HfOx film inserted in Zener diode cell could induce switching and synaptic behaviors. This design could simplify the functional memory arrays as the building block of AI hardware.

Our recent article on #Neuromorphic Spike Freq. Adaptation published in Nature Portfolio Comm. Eng. Thanks co-authors Sai Sukruth Bezugam Melika Payvand Elisabeth Sounak & Chittotosh. FITT IIT Delhi, UC Santa Barbara ETH Zurich TCS Research rdcu.be/dx1Yg

Researchers from UT Austin report an ultra-fast #memristor based on thin sheets of hexagonal Boron Nitride exhibiting the shortest observed switching #speed (120 ps) among #2D #memristors and low switching #energy (2pJ). Now out👉Nature Communications nature.com/articles/s4146…

Inspired by the human cochlea, Milozzi et al. NEDL - NanoElectronic Device Lab develop #neuromorphic #circuits for #memristive tonotopic mapping via #volatile resistive switching #memory devices. Now out 👉Nature Communications nature.com/articles/s4146…

This IEEE Spectrum article written by Mark Liu (TSMC chairman) and my colleague, Philip Wong (H.-S. Philip Wong), is required reading for anyone working on semiconductors and AI today. spectrum.ieee.org/trillion-trans…

Researchers from Sandia National Labs, Seoul National University and Texas A&M University experimentally demonstrate that tunable #stochastic #memristors can perform #cryptographic key generation and #energyefficient #encryption and #computing. Now out 👉Nature Communications nature.com/articles/s4146…

Wang et al. report a nanoscale #LED with memory #electroluminescence, which can realize multiple action potentials combinations through a single optical channel in bio-inspired #afferent #nerves. Now out 👉 Nature Communications nature.com/articles/s4146…

Prof. Duygu Kuzum and colleagues from UC San Diego report a filament-free trilayer metal-oxide bulk switching #RRAM technology for #edge #computing for an #autonomous #navigation task. Now out Nature Communications 👉 nature.com/articles/s4146…

"Neurons are noisy" is one of the bigger (and for neural computing, detrimental...) misconceptions around the brain. Everything about brains' development and anatomy is so intricate, why would communication be random and unpredictable? Any variation is there for a reason.

Check out "DenRAM", the first implementation of delay-based dendritic architecture using #RRAM+CMOS technology, which we show is a natural match for efficient temporal signal processing Nature Communications. paper: nature.com/articles/s4146… code: github.com/EIS-Hub/DenRAM Thread 🧵

New paper in Nature Electronics. The first ever demonstration of a scalable high temperature non-volatile memory technology. Great collaboration with Troy Olsson, Eric Stach and Air Force Research Lab - AFRL. Penn Penn Engineering Penn Electrical and Systems Engineering Materials Science and Engineering at Penn Nano nature.com/articles/s4192…

Researchers from Fudan University develop bio-inspired firing feature-driven #neuralcircuits with scalable #memristive neurons for #robotic obstacle avoidance. Now out👉Nature Communications nature.com/articles/s4146…

Most low-temperature fabricated devices are not CMOS compatible because they wont survive back-end thermal process which induce element intermixing, damaging interfaces. In this paper, we intentionally exploit the intermixing to make better devices for #AI applications!

Teacher’s proud moment. My PhD student won best paper award!

Another paper has been published led by my MSc students. We propose a new statistical metric “Fluctuation Factor (FF)” to predict how likely the circuit is to produce good random bits by quantifying variations in the switching process, which other commonly used metrics can’t.