CERN Computing

New technologies to boost real-time AI capabilities for the ATLAS experiment! At the High-Luminosity LHC (#HiLumiLHC), ATLAS will face unprecedented challenges in terms of data rates and the complexity of the collision events. The ATLAS Hardware Trigger system must decide, in real time, which events are potentially interesting enough to keep — and it must do so within microsecond-scale latency budgets. Featured here is a prototype board for the ATLAS Global Trigger system. At the heart of this special piece of hardware is an AMD VP2802 FPGA – the large chip in the middle of the picture. It is one of the most powerful FPGAs in the market and includes novel, super-fast, digital processors called AI Engines alongside programmable logic. Using AI Engines opens up a new frontier for ATLAS: - integrating more sophisticated machine-learning algorithms directly inside the real-time trigger path, - while respecting the extreme constraints on latency, throughput and power. This will enhance the ATLAS Experiment’s ability to explore new phenomena at the energy frontier and pursue rare processes that could shed light on the universe’s deepest mysteries. This development was supported by the #NextGenTriggers project, as part of its work on advanced real-time processing technologies for the HL-LHC.

🏃♂️➡️ 🏃♀️➡️ 📣 Will you be our next colleague? We are already looking forward to meeting you! 🤗 ------ DevOps/Kubernetes Web Hosting Engineer 24 month contract Hybrid https://lnkd.in/eHNg5KsM Software and DevOps Engineer 24 month contract Hybrid https://lnkd.in/eEzSsSVs Systems & Security Engineer 24 month contract Hybrid https://lnkd.in/eZybMjgP DevOps Engineer - WordPress Platform 24 month contract Hybrid https://lnkd.in/e4zVtDkc Full Stack Software Engineer and Service Manager 60 month contract Hybrid https://lnkd.in/ejCgrrsk Ceph Software Engineer 60 month contract Hybrid https://lnkd.in/eS3rEaHE Data Analytics Infrastructure Engineer 36 month contract Hybrid https://lnkd.in/ek_XVxMd Database Platform Engineer 24 month contract Hybrid https://lnkd.in/exZR-8vN

What do gravitational waves, real-time data processing and AI at the edge have in common? From 5 to 8 May 2026, CERN hosted the AI for Gravitational Waves Workshop, bringing together more than 80 participants in person and online to explore how #AI and #MachineLearning are reshaping gravitational-wave science. With the support of #NextGenTriggers, the workshop covered topics ranging from low-latency detection and parameter inference to fast waveform generation, streaming ML pipelines, detector monitoring, predictive maintenance and accelerated computing with GPUs and FPGAs. 👇 Read more about it here: https://lnkd.in/e2Dwx5Zh 🏆And congratulations to Ludovica Carbone and Lucia Papalini, winners of the Best Poster Award and Best Presentation Award respectively!

Computing power is the often invisible engine that enables scientific research. As CERN prepares for the High-Luminosity Large Hadron Collider (HL-LHC) era, the data centres that will support research in the coming years are undergoing a quiet but crucial transformation. ✅ The Meyrin Data Centre (MDC) is CERN's primary hub for data storage and critical applications. Here, work is underway to strengthen something most people never think about: the floor. Indeed, modern computing hardware is becoming denser, more powerful, and significantly heavier! ✅ The Prévessin Data Centre (PDC) is stepping into an even bigger role as a computing hub and backup site. After two successful years of operation in its first phase, attention is now turning to expansion. Near-term improvements include activating a heat recovery system and enhancing cooling performance to reduce water usage—both key steps toward greater #sustainability. ✅ Equally important is the #network that supports all data transfers from all experimental sites distributed across the CERN campus. Starting in June 2026, a major fibre upgrade will modernise the MDC’s connectivity, ready to handle the enormous surge in transferred data volumes expected from HL-LHC  experiments. 🌟 Looking beyond 2026, the MDC is set for a broader evolution. Plans include: simplifying electrical systems and improving energy efficiency, introducing dual power feeds for greater reliability, upgrading racks to accommodate modern, larger hardware, replacing aging power components, transitioning to a more flexible, modular backup power architecture, exploring advanced cooling solutions to reduce energy consumption. The Prévessin Data Centre (PDC) will dramatically increase its computing capacity, from the current 4 MW (megawatts) of capacity to the future 12 MW.   As the HL-LHC era approaches, the CERN data centres are being reshaped to meet the intense demands of next-generation science while remaining safe and reliable for decades to come. #HighLumiLHC

At the High-Luminosity LHC, the challenge will not only be producing more data, but making smarter decisions about which data to keep. Combining AI and machine learning with specialised hardware such as FPGAs, and tools like #hls4ml, the #NextGenTriggers project aims to help experiments extract more physics information from extreme data rates and improve their sensitivity to ultra-rare phenomena. Read more in this article 👇

Applications are now open for the 17th edition of the ESC - Efficient Scientific Computing School - INFN! Organised by INFN in partnership with the CERN #NextGenTriggers project, the school will take place from 12 to 22 October 2026 in the ancient fortress of Bertinoro, Italy. The programme is designed for PhD students and young researchers working on the development of scientific applications and software tools for research. Participants will explore architectures, tools and methodologies for large-scale scientific computing applications, with lectures covering: 🔹 Current trends in processor architectures 🔹 Parallel programming 🔹 Advanced C++ 🔹 Efficient memory use 🔹 Floating-point computation 🔹 Heterogeneous programming environments 🔹 Multi-threaded computing, GPUs and MPI-based cluster computing For more information and registration: https://lnkd.in/e9MQhvDM Anna Dalla Bona

The Web: invented to serve scientists, now woven into the fabric of society. What began as a tool for scientific collaboration has evolved into the backbone of how we communicate, learn, work, and create everyday. It has democratized access to information and has reshaped society. But with that transformation comes responsibility. Let’s feed the Web with content that informs, connects, and empowers—#responsibly.

🏃♀️➡️ 🏃♂️➡️ 🔜 One week to application deadline. Be smart, apply today! https://lnkd.in/exvH-iYM

#REANA is the CERN-developed platform that allows scientists to run data analysis pipelines on remote compute clouds. It enables near-data computation—sending computational workflows to where large scientific datasets are stored, rather than transferring massive volumes of data to the researcher. We have developed it for use in particle physics—a field that generates enormous data volumes—but it is applicable to many other domains, including astronomy and life sciences. Thanks to REANA, researchers can execute their analyses directly at the data source and can manage containerized workflows across federated computing resources. REANA was recently featured as a EOSC Association use case. CERN manages one of EOSC nodes and, besides REANA, it shares with the EOSC Federation other in-house developed software solutions, including #Zenodo, #Indico, #InvenioRDM, as well as #LHC data through the Open Data portal. Find out more about Open Science at CERN 👇 : https://openscience.cern/ A recent article summarizing many initiatives CERN is involved in 👇 : https://lnkd.in/eXr98FVA ------- #openscience