I’m an AI Research Scientist within the Neuromorphic Computing Lab at Intel, where we develop the Intel Loihi 2 accelerator. I’m also a CS PhD student at LMU Munich, advised by Prof. Eyke Hüllermeier at the Chair of AI and ML.
My research interests can be broadly grouped into:
- Hardware-software co-design for efficient ML
- Novel recurrent architectures
- LLMs and emerging capabilities
Feel free to reach me on at @lessandroPierro or by email at alessandropierro [at] duck [dot] com. Happy to jump on a call or grab a coffee.
If you are interested, here is my Curriculum Vitae.
News
-
July 2025: Release of SparseRNNs
Our JAX codebase to train sparse and quantized SSMs has been released at IntelLabs/SparseRNNs. -
July 2025: Invited talk at the Telluride Neuromorphic Cognition Engineering Workshop
I’m spending three weeks at the workshop in Telluride, CO. I’ll be presenting my work on SSMs and collaborate on bringing LLMs to neuromorphic chips. -
May 2025: Paper accepted at ICML 2025
Our work demonstrates unstructured sparsity (in both weights and activations) to enable more efficient inference for state space models at the edge. We validate this on an audio denoising task running in real-time on Intel Loihi 2. -
May 2025: Started new role as AI Research Scientist at Intel
-
February 2025: Paper published on Nature Communications
The work presents NeuroBench, a large community effort led by Jason to establish collaborative, fair, and representative benchmarks for neuromorphic computing. I was glad to contribute the benchmarking methodology and results from my M.Sc. thesis to the QUBO task in the system track. -
February 2025: Invited talk at Loughborough University
Juan Sebastian Totero Gongora invited me to present my work on Intel Loihi 2 at the Lindau seminar and toured me around their photonics lab, brainstorming applications of neuromorphic processors to high-frequency control. -
October 2024: Invited talk at Ostfold University College
Stefano Nichele invited me to present my work on Loihi 2 at the Neuromorphic Computing: Problems and Prospects workshop. -
July 2024: Presented workshop paper accepted at ICML
Steve and I attended ICML in Vienna and presented two workshop papers on quantization for S5 and Mamba. -
May 2024: Invited talk at Glasgow University
Giulia Marcucci invited me to present my work at the ABI Salon interdisciplinary seminar.
Selected Publications
For a complete list of my publications, you can refer to Google Scholar.
- Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsity
Alessandro Pierro*, Steven Abreu*, Jonathan, Timcheck, Philipp Stratmann, Andreas Wild, Sumit Bam Shrestha
ICML 2025 - The Neurobench Framework for Benchmarking Neuromorphic Computing Algorithms and Systems
Jason Yik, Korneel Van den Berghe, Douwe den Blanken, Younes Bouhadjar, Maxime Fabre, Paul Hueber, Weijie Ke, Mina A. Khoei, Denis Kleyko, Noah Pacik-Nelson, Alessandro Pierro, et al.
Nature Communications, 2025 - Solving QUBO on the Loihi 2 Neuromorphic Processor
Alessandro Pierro*, Philipp Stratmann*, Gabriel Fonseca Guerra*, Sumedh Risbud, Timothy Shea, Ashish Rao Mangalore, Andreas Wild
arXiv, 2024 - Mamba-PTQ: Outlier Channels in Recurrent Large Language Models
Alessandro Pierro*, Steven Abreu*
ICML 2024 2nd Workshop on Efficient Systems for Foundation Models - Q-S5: Towards Quantized State Space Models
Steven Abreu*, Jens E. Pedersen*, Kade M. Heckel*, Alessandro Pierro
ICML 2024 Workshop on Next-Generation Sequence Modeling Architectures