Applied Physicist at CERN, CMS Experiment, Level 1 Trigger, Fast Machine Learning
Data taking at the LHC recently ended , and we now begin a period of several years of upgrades to the detector and trigger system. Before we can install the new systems, we need to clear out the legacy equipment that won’t take part in the Phase 2 upgrade. Today I had the bittersweet experience of decommissioning part of the CMS trigger system, together with Aaron Bundock and Sudan Paramesvaran .…
Last week I presented “AI at the edge: underground and in space” at the AI for Gravitational Waves Workshop. The workshop was held at CERN from 5-8 May and brought together experts and practitioners in AI from the different gravitational wave experiments as well as some perspectives on other AI activities going on at CERN. I presented how we’re using edge AI technologies at the LHC experiments and…
Setting the scene Recently I was taking part in some debugging of an ML model deployment for the CMS Level 1 Trigger in Run 3. The Level 1 Trigger is the tier of processing at the CMS experiment. It’s responsible for computing which particles and other physics quantities can be seen in an LHC collision, and making a decision whether that particular collision is worthy of being read out from the…
I was happy to be invited to present an overview of AI projects with industry relevance at the CERN Openlab Technical Workshop 2026. Openlab provides a structured and collaborative framework for industry and research organisations to engage with CERN researchers. In my work on Fast Machine Learning, as a developer of tools like hls4ml and conifer , and as a practitioner at the CMS experiment, I’ve…
In my daily work on FPGAs for the CMS Level 1 Trigger we usually are not limited by I/O into the devices. We work with boards, like the Serenity and APx, which can support 4-5 Tb/s input on around 100 optical fibres with very low latency - 100 ns or so for a hop from one board to the next. This is very convenient for designing algorithms, since often all (or at least many) of the inputs can be…
I was delighted to participate in the CERN for Climate Action Day event, organised by CIPEA, this October 2025. CIPEA is the CERN Innovation Programme on Environmental Applications, under whose umbrella falls the Edge SpAIce project that I lead at CERN. During this half-day event, representatives from many of the projects under the CIPEA banner presented their work. I was particularly impressed…
I was delighted to be invited to speak at a side event of the UN Plastic Treaty INC-5.2 at the CERN IdeaSquare. The excellent event was organized by the Polarquest Association, with talks from experts in citizen science, marine biology, and geography. I presented the role of AI in detecting plastics pollution in the oceans with the Edge SpAIce project. It was a great pleasure to present, and to…
We recently put out our first public result for the CMS NextGen Triggers “Practical Real Time AI” task on improved jet tagging. The project started all the way back in late 2023 as the winning hackathon project at the CMS ML@L1T workshop (read more about that workshop in this post ). Incorporating recent best practices from the offline world, we used a large sample of different processes and a…
</a> I was very happy to teach for a day and a half on High Level Synthesis (HLS) at Milano Bicocca as part of a week long FPGA school. The overall course organisation was fantastic, Simone did a great job bringing the school together, Mirko and Andrea put together excellent course material, and Francesco and Giulio were tirelessly on hand during the exercises. In the HLS part, I focussed on two…
Some members of my group at CERN had the great idea recently to start a weekly coffee morning before our group meetings. For me this was an irresistable invitation to bring high-effort manual espresso to the work place. Chris and I hosted, with half of each of our home set up. Chris provided the grinder and I brought the espresso machine. The machine itself is a Flair 58 lever espresso machine -…