Rapid learning with phase-change memory-based in-memory computing through learning-to-learn

T Ortner, H Petschenig, A Vasilopoulos… - Nature …, 2025‏ - nature.com
There is a growing demand for low-power, autonomously learning artificial intelligence (AI)
systems that can be applied at the edge and rapidly adapt to the specific situation at …

The inherent adversarial robustness of analog in-memory computing

C Lammie, J Büchel, A Vasilopoulos… - Nature …, 2025‏ - nature.com
A key challenge for deep neural network algorithms is their vulnerability to adversarial
attacks. Inherently non-deterministic compute substrates, such as those based on analog in …

Efficient scaling of large language models with mixture of experts and 3D analog in-memory computing

J Büchel, A Vasilopoulos, WA Simon, I Boybat… - Nature Computational …, 2025‏ - nature.com
Large language models (LLMs), with their remarkable generative capacities, have greatly
impacted a range of fields, but they face scalability challenges due to their large parameter …

Kernel approximation using analogue in-memory computing

J Büchel, G Camposampiero, A Vasilopoulos… - Nature Machine …, 2024‏ - nature.com
Kernel functions are vital ingredients of several machine learning (ML) algorithms but often
incur substantial memory and computational costs. We introduce an approach to kernel …

Programming weights to analog in-memory computing cores by direct minimization of the matrix-vector multiplication error

J Büchel, A Vasilopoulos, B Kersting… - IEEE Journal on …, 2023‏ - ieeexplore.ieee.org
Accurate programming of non-volatile memory (NVM) devices in analog in-memory
computing (AIMC) cores is critical to achieve high matrix-vector multiplication (MVM) …

Improving the accuracy of analog-based in-memory computing accelerators post-training

C Lammie, A Vasilopoulos, J Büchel… - … on Circuits and …, 2024‏ - ieeexplore.ieee.org
Analog-Based In-Memory Computing (AIMC) inference accelerators can be used to
efficiently execute Deep Neural Network (DNN) inference workloads. However, to mitigate …

Demonstration of 4-quadrant analog in-memory matrix multiplication in a single modulation

M Le Gallo, O Hrynkevych, B Kersting… - npj Unconventional …, 2024‏ - nature.com
Analog in-memory computing (AIMC) leverages the inherent physical characteristics of
resistive memory devices to execute computational operations, notably matrix-vector …

Learning-to-learn enables rapid learning with phase-change memory-based in-memory computing

T Ortner, H Petschenig, A Vasilopoulos… - arxiv preprint arxiv …, 2024‏ - arxiv.org
There is a growing demand for low-power, autonomously learning artificial intelligence (AI)
systems that can be applied at the edge and rapidly adapt to the specific situation at …

Opposing Mean Error Compensation for Accuracy Enhancement in Analog Compute-in-Memory With Resistive Switching Devices

WH Huang, W Lee, JS Kim, WT Koo… - … on Electron Devices, 2024‏ - ieeexplore.ieee.org
Analog compute-in-memory (ACiM) systems show promise for energy-efficient AI inference,
but their performance is hindered by variations in conductance, resulting in reduced …

Exploring learning techniques for edge AI taking advantage of non-volatile memories

M Martemucci - 2024‏ - theses.hal.science
Learning-capable edge artificial intelligence (AI) systems require both inference and
learning capabilities combined with energy efficiency. However, no existing memory …