I am CTO at Physical Superintelligence, where I am building and leading the company's research and engineering organization across physics research, AI research, and engineering. I joined PSI early to build that organization and contributed to the company's $58M seed round, led by Breakthrough Energy Ventures.
PSI is a public benefit corporation building an AI-native physics lab. Human scientists direct research campaigns; virtual physicists carry them out, from hypothesis through simulation, verification, and engineering. A result counts as a discovery when independent verifications converge on it. The first proving ground is the physical infrastructure of AI: data centers and the factories that build them.
The organization I lead has to do three things at once: produce physics results that survive verification, build the AI systems that produce them at scale, and ship engineering that customers can run. That is why physics, AI research, and engineering sit in one organization rather than three.
Bio
Before PSI I worked at NVIDIA, IBM Research, and the Department of Energy's national laboratories, on the systems side of AI.
At NVIDIA I led agentic GPU kernel development and optimization for frontier AI lab workloads and set the performance evaluation strategy for CuTile, NVIDIA's next-generation GPU programming model.
At IBM Research I led IBM's enterprise generative AI coding assistant from research to production: an 8-person team inside a 60+ person effort across IBM and Red Hat. Earlier I led the HPC systems team on Summit and Sierra, the two fastest supercomputers in the world at launch, supporting large-scale scientific simulation at DOE laboratories. During that period I created NYU Courant's first graduate course on high performance machine learning, still taught today.
At PNNL I built distributed runtime systems that spun off into Trovares, now RocketGraph.
PhD in Computer Architecture, UPC Barcelona. IPDPS Best Paper Award (2012). 30+ publications, 18 patents filed, 12 granted.
Education
- Ph.D. in Computer Architecture — Polytechnic University of Catalunya (UPC), Barcelona
- M.Sc. in Computer Engineering — University of Rome Tor Vergata, Rome
- B.Sc. in Computer Engineering — Roma Tre University, Rome
Featured Work
Research and engineering organization
Physics research, AI research, and engineering under one technical mandate. Autonomous campaigns from hypothesis to verified, deployed systems.
CuTile Programming Model
AI-driven GPU kernel development and next-generation programming model for production workloads.
Watson Code Assistant
Led creation of the first generative model for IBM's AI-powered code assistant product.
Summit & Sierra Supercomputers
System software for the world's #1 and #2 fastest supercomputers (2018 TOP500).
Awards
- IBM Outstanding Technical Achievement Award (2023), for leading the creation of the first generative model for Watson Code Assistant
- IBM Research Division Award (2018), for the CORAL Summit and Sierra supercomputers
- HPCwire Readers' and Editors' Choice, Top Supercomputing Achievement (2018), Summit and Sierra (awarded to ORNL and LLNL; IBM system software team)
- PNNL Outstanding Performance Award (2014), for the GEMS graph engine
- IPDPS Best Paper Award (2012), for Evaluating the Impact of TLB Misses on Future HPC Systems
- HiPEAC Collaboration Grant (2010), for research at IBM T. J. Watson Research Center
In the Press
Physical Superintelligence (2026)
Announcing Physical Superintelligence, a public benefit corporation pointing superintelligence at the hardest problems in physics: producing new physics, at scale, discovered and commercialized safely and verifiably for broad public benefit.
- Physical Superintelligence: the AI-native physics lab
- Physical Superintelligence, Dr. Alex Wissner-Gross, The Innermost Loop
CUDA Tile Programming Model (2025)
- NVIDIA Introduces CUDA 13.1 with CUDA Tile — InsideHPC
- NVIDIA CUDA 13.1 Powers Next-Gen GPU Programming — NVIDIA Developer Blog
AI for Code / Watson Code Assistant (2022–2023)
Summit & Sierra Supercomputers (2018)
- Two DOE Supercomputers Top List of World's Fastest — U.S. Department of Energy
- Sierra Honored as Top Supercomputing Achievement — LLNL / HPCwire
- Summit Supercomputer Is Already Making Its Mark on Science — HPCwire
Graph Analytics (2014–2019)
Selected Papers
For a full list, see Google Scholar.
AI and Machine Learning
- Towards Learning (Dis)-Similarity of Source Code from Program Contrasts
ACL 2022 - VELVET: a noVel Ensemble Learning approach to automatically locate VulnErable sTatements
SANER 2022 - D2A: A Dataset Built for AI-Based Vulnerability Detection Methods Using Differential Analysis
ICSE-SEIP 2021 - Probing Model Signal-Awareness via Prediction-Preserving Input Minimization
ESEC/FSE 2021 - Exploring Software Naturalness through Neural Language Models
ArXiv Preprint, 2020
High Performance Computing and Systems
- In-Memory Graph Databases for Web-Scale Data
IEEE Computer, 2015 - Scaling Semantic Graph Databases in Size and Performance
IEEE Micro, 2014 - Scaling Irregular Applications through Data Aggregation and Software Multithreading
IPDPS 2014 — Best Paper Award - Evaluating the Impact of TLB Misses on Future HPC Systems
IPDPS 2012 - A Quantitative Analysis of OS Noise
IPDPS 2011
Writing
On technology
Superintelligent Design
AI design will be beyond our imagination once we automate validation. In the physical world a wrong answer costs more, so the models AI designs with have to be checked far better than they are today.
Read on LinkedIn →Converting Compute Into Breakthroughs
For fifty years, software companies sold tools and left customers to turn them into outcomes. That arrangement has shaped the industry for decades — and it's now coming apart.
Read on LinkedIn →Tuning Flash Attention for Peak Performance in NVIDIA CUDA Tile
A deep dive into optimizing Flash Attention on Blackwell GPUs using cuTile, achieving 1.60-1.66x speedups through loop splitting, block remapping, and autotuning.
Read on NVIDIA Developer Blog →Generation Is Cheap. Verification Is Not.
Why complex software will survive the AI shock. The hard part was never writing the code — it was knowing whether the code was right.
Read on LinkedIn →The Physics of AI Systems: Four Fundamental Laws
AI systems have their own physics. Not metaphorical: actual constraints rooted in mathematics and the physical properties of silicon, light, and electrons.
Read on LinkedIn →Philosophy
Get in Touch
Hiring: PSI is hiring physicists, AI researchers, and engineers in Boston and San Francisco. Open roles
Partnerships, media, investors: partnerships@psi.inc
Me: LinkedIn