Design of Analog-AI Hardware Accelerators for Transformer-based Language Models (Invited)
- Geoffrey Burr
- Sidney Tsai
- et al.
- 2023
- IEDM 2023
Manuel Le Gallo-Bourdeau
Title
Researcher in Neuromorphic and In-Memory Computing
Bio
Manuel Le Gallo is a Staff Research Scientist in the In-Memory Computing group.
His main research interest is in using phase-change memory (PCM) devices for non-von-Neumann computing. He is a member of the team that has created randomly spiking neurons using phase-change materials that store data and perform computations like the human brain. This is a significant step forward in the development of energy-efficient, ultra-dense, integrated neuromorphic computing technologies.
He originally joined IBM Research - Zurich as a Master's student in 2013, working on the electrical characterization of PCM cells and the experimental investigation of melting/crystallization and subthreshold conduction. He conducted his PhD at IBM affiliated with the Institute of Neuroinformatics at ETH Zurich.
He was appointed IBM Master Inventor in 2019 and is a recipient of the MIT Technology Review's 2020 Innovators Under 35 award.
Manuel holds a doctoral degree in Electrical Engineering (DrSc) from ETH Zurich, a Master's degree in Electrical Engineering and Information Technology (MSc) from ETH Zurich, an undergraduate degree of Cycle Ingénieur Polytechnicien from Ecole Polytechnique (l'X), Palaiseau, France, as well as a Bachelor's degree in Engineering Physics (BEng) from Ecole Polytechnique de Montréal, Canada.
Publications
In-Memory Compute Chips with Carbon-based Projected Phase-Change Memory Devices
- G.S. Syed
- K. Brew
- et al.
- 2023
- IEDM 2023
Programming Weights to Analog In-Memory Computing Cores by Direct Minimization of the Matrix-Vector Multiplication Error
- 2023
- IEEE JESTCS
Using the IBM Analog In-Memory Hardware Acceleration Kit for Neural Network Training and Inference
- 2023
- APL Mach. Learn.
Deep neural network inference with a 64-core in-memory compute chip based on phase-change memory
- 2023
- E\PCOS 2023
Hardware-aware training for large-scale and diverse deep learning inference workloads using in-memory computing-based accelerators
- Malte J. Rasch
- Charles Mackin
- et al.
- 2023
- Nature Communications
Deep neural network inference with a 64-core in-memory compute chip based on phase-change memory
- 2023
- NVMTS 2023
A 64-core mixed-signal in-memory compute chip based on phase-change memory for deep neural network inference
- Manuel Le Gallo
- Riduan Khaddam-Aljameh
- et al.
- 2023
- Nature Electronics
AnalogNAS: A Neural Network Design Framework for Accurate Inference with Analog In-Memory Computing
- Hadjer Benmeziane
- Corey Lammie
- et al.
- 2023
- EDGE 2023
In-memory computing for accelerating deep neural networks and neuro-vector-symbolic architectures
- 2023
- AICAS 2023
Patents
- US
- 10248323
- US
- 10217046
- US
- 10210138
- US
- 10114613
- US
- 10079058
- US
- 9996793
- US
- 9990580
- US
- 9910639
- US
- 9785885
- US
- 9767408
Projects
Blog posts
IBM Research's latest analog AI chip for deep learning inference
NewsAbu Sebastian, Manuel Le Gallo-Bourdeau, and Vijay NarayananThe hardware behind analog AI
ExplainerManuel Le Gallo-BourdeauNovel synaptic architecture for brain inspired computing
ResearchIrem BoybatIBM scientists demonstrate mixed-precision in-memory computing for the first time; hybrid design for AI hardware
Research- See more blog posts
Top collaborators
AS
Abu Sebastian
Distinguished Scientist & Manager
CL
AV
KE
Kaoutar El Maghraoui
Principal Research Scientist and Manager, AIU Spyre Model Enablement, AI Hardware Center