The MIT AI Hardware Program is an academia-industry initiative between the MIT School of Engineering and MIT Schwarzman College of Computing. We work with industry to define and bootstrap the development of translational technologies in hardware and software for the AI and quantum age.

Our annual symposium includes a keynote talk from professor Elsa Olivetti, reviews of the current project portfolio, presentations on new projects, and a networking reception featuring a poster session and interactive demos.

10:00 - 10:10

Year in Review & the Year Ahead

Program Co-Leads

Jesús del Alamo, Donner Professor; Professor, Electrical Engineering and Computer Science; MacVicar Faculty Fellow
Aude Oliva, Director of Strategic Industry Engagement, MIT Schwarzman College of Computing; CSAIL Senior Research Scientist

10:10 – 11:30

Project Reviews

A 14-nm Energy-Efficient and Reconfigurable Analog Current-Domain In-Memory Compute SRAM Accelerator
Aya Amer, Postdoctoral Associate, Research Laboratory of Electronics

Wafer-Scale 2D Transition Metal Dichalcogenides for Neuromorphic Applications
Jiadi Zhu, PhD Candidate, Electrical Engineering and Computer Science

Increasing Architectural Resilience to Small Delay Faults
Peter Deutsch and Vincent Ulitzsch, PhD Candidates, Electrical Engineering and Computer Science

CIRCUIT: A Benchmark for Circuit Interpretation and Reasoning Capabilities of LLMs
Yan Xu, Ph.D. Candidate, Electrical Engineering and Computer Science

Efficient Large Language Models and Generative AI
Song Han, Associate Professor, Electrical Engineering and Computer Science

11:40 – 12:00

UROP (Undergraduate Research Oppurtunites Program) Pitches

PERE-Chains: AI-Supported Discovery of Privilege Escalation and Remote Exploit Chains
Cristián Colón, Undergraduate, Engineering and Computer Science

Computing with Heat
Caio Silva, Undergraduate, Physics

Simulation of Optical Phase Change Modulator for Analog Photonic Applications
Anthony Donegan, Undergraduate

Ferroelectric Memory Devices for AI Hardware
Tyra Espedal, Undergraduate, Physics

1:00 – 1:30

Keynote

The Climate and Sustainability Implications of Generative AI
Elsa Olivetti, Professor, Materials Science and Engineering

1:30 – 2:30

Highlights: Prospective New Projects

Hardware-efficient Neural Architectures for Language Modeling
Lucas Torroba Hennigen, PhD Candidate, Computer Science and Artificial Intelligence Laboratory

Ferroelectric AI Hardware: Overcoming Conventional Paradigms and Scalability Limits
Suraj Cheema, Assistant Professor, Materials Science and Engineering & Electrical Engineering and Computer Science

Analog Computing with Inverse-Designed Metastructures
Giuseppe Romano, Research Scientist, Institute for Soldier Nanotechnologies

Magnetic Tunnel Junction for Stochastic Neuromorphic Computing
Luqiao Liu, Associate Professor, Electrical Engineering and Computer Science

Compressing (in) the Wild: Continual Fine-tuning of Autoencoders for Camera Trap Image Compression
Timm Haucke, PhD Candidate, Electrical Engineering and Computer Science

Declarative Optimization for AI Workloads
Michael Cafarella, Research Scientist, Computer Science and Artificial Intelligence Laboratory (CSAIL)

2:30 – 3:30

Research Showcase

A Unified Framework for Sparse Plus Low-Rank Matrix Decomposition for LLMs
Mehdi Makni, PhD Candidate, MIT Operations Research Center

Hardware-Aware Algorithms for Large Neural Network Compression
Xiang Meng and Ryan Lucas, PhD Candidates, MIT Operations Research Center

Extracting Cellular Automaton Rules from Biological Systems with Generative Pretrained Transformers
Jaime Berkovich, PhD Candidate, Materials Science and Engineering

How Much Potential is There for Further GPU Progress?
Emanuele Del Sozzo, Research Scientist, Computer Science & Artificial Intelligence Lab

In-Memory Sparsity: Enable Efficient Unstructured Element-Wise Sparse Training from the Bottom
Hongkai Ning, Postdoctoral Associate, Research Laboratory of Electronics

Learning to Keep a Promise: Scaling Language Model Decoding Parallelism with Learned Asynchronous Decoding
Tian Jin, PhD Candidate, Electrical Engineering and Computer Science

SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models
Muyang Li and Yujun Lin, PhD Candidates, Electrical Engineering and Computer Science

Computing with Heat
Caio Silva, Undergraduate, Physics