📢 Final #EPICURE webinar on profiling! Matteo Murdaca will introduce #AMD tools for #profiling and optimising applications on @lumi-supercomputer.eu. 📅 29 September 2026 | 10:00 CET | Online 🔗 epicure-hpc.eu/2026/05/22/w... #EPICURE #HPC #Profiling @eurohpc-ju.bsky.social

🚀 Registration is open for the next EPICURE Hackathon! Join us on 7–9 October 2026 in Kraków or online to optimise your HPC and AI workloads with expert support. Register 👉 epicure-hpc.eu/2026/07/16/e... #EPICURE #HPC #AI @eurohpc-ju.bsky.social

🎥 Did you miss our webinar on NVIDIA Nsight Systems for heterogeneous HPC platforms? Our webinar, “From traces to optimisation with NVIDIA Nsight Systems on heterogeneous HPC platforms”, featuring Laura Bellentani (CINECA), is available online. ▶️ www.youtube.com/watch?v=-o7z...

Webinar | From traces to optimisation with NVIDIA Nsight Systems on heterogeneous HPC platforms

YouTube video by EPICURE Project

youtube.com

🚀 Up to 8× faster #AI performance on #AMD GPUs #EPICURE and Qubit Pharmaceuticals optimised molecular dynamics simulations on @LUMIhpc using #JAX, reproducible containers and profiling. 🔗 Blog Post: epicure-hpc.eu/2026/07/09/g... 🔗 Success Story: epicure-hpc.eu/2026/07/28/2... #HPC

🎥 How do you support #HPC users ranging from complete beginners to seasoned experts? We spoke with David Vicente (@bsc-cns.bsky.social) about coordinating code porting and scaling support across @eurohpc-ju.bsky.social centres. Watch the full interview👇 www.youtube.com/watch?v=Q8-n... #EPICURE

How EPICURE helps researchers scale their codes faster | WP2 with David Vicente

YouTube video by EPICURE Project

youtube.com

⚡ 𝘠𝘰𝘶𝘳 𝘤𝘰𝘥𝘦 𝘤𝘰𝘶𝘭𝘥 𝘣𝘦 𝘧𝘢𝘴𝘵𝘦𝘳. 𝘓𝘦𝘵'𝘴 𝘧𝘪𝘹 𝘵𝘩𝘢𝘵. 📅 SAVE THE DATE | EPICURE Hackathon: Code Optimisation for HPC & AI Workloads 🗓️ 7–9 Oct 2026 📍 ACC Cyfronet AGH, Kraków 🇵🇱 & Online Details soon 👉 shorturl.at/W5K8A #EPICURE #HPC #AI #EuroHPC @eurohpc-ju.bsky.social @inesctec.bsky.social

⏰ 𝗝𝗼𝗶𝗻 𝘁𝗼𝗺𝗼𝗿𝗿𝗼𝘄’𝘀 𝘄𝗲𝗯𝗶𝗻𝗮𝗿 to learn how #NVIDIA Nsight Systems can help you trace and profile #CPU - #GPU applications, identify performance bottlenecks, and make informed optimisation decisions for heterogeneous #HPC workloads. 📅 Tomorrow 🔗 shorturl.at/ddKsw @cineca.bsky.social

🚀 Optimising application performance starts with understanding where time is really spent. Discover how Frédéric Pinel (@luxprovide.bsky.social) uses #MAQAO to identify bottlenecks, analyse performance and apply targeted optimisations for #HPC applications. 🔗 shorturl.at/VVb57

⏱️ 𝗬𝗼𝘂𝗿 𝗮𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗿𝘂𝗻𝘀 𝗼𝗻 𝗯𝗼𝘁𝗵 𝗖𝗣𝗨𝘀 𝗮𝗻𝗱 𝗚𝗣𝗨𝘀. 𝗕𝘂𝘁 𝗱𝗼 𝘆𝗼𝘂 𝗸𝗻𝗼𝘄 𝘄𝗵𝗲𝗿𝗲 𝗶𝘁’𝘀 𝘀𝗽𝗲𝗻𝗱𝗶𝗻𝗴 𝗶𝘁𝘀 𝘁𝗶𝗺𝗲? The next webinar explores how #NVIDIA Nsight Systems can help you understand application behaviour across heterogeneous #HPC platforms. 📅 July 10 🔗 shorturl.at/ddKsw @cineca.bsky.social

🎥 Advanced technical support, community building, or knowledge exchange: which matters most for the future of #HPC? We asked Janne Ignatius, #EPICURE Coordinator, to share his thoughts on the challenges and opportunities shaping support for #EuroHPC users👇 @csc.fi @eurohpc-ju.bsky.social

EPICURE’s vision for the future of HPC support in Europe

YouTube video by EPICURE Project

youtu.be

That's a wrap on #ISC26! Thanks to everyone who connected with the #EPICURE team! Great conversations on supporting users across the European #HPC ecosystem. See you next year! 🤩 @eurohpc-ju.bsky.social @inesctec.bsky.social @uminho.bsky.social #EuroHPC #ISC26 #HPC

🧬 Faster #genome analysis for large-scale research With EPICURE support, the Latvian genome reference project ran 3× faster on #MeluXina with #EPICURE #support, a boost for the Genome of Europe. 🔗 shorturl.at/oaHGr @luxprovide.bsky.social @eurohpc-ju.bsky.social

🎥 The recording of our webinar "POP CoE performance metrics: a beginner’s guide with Score-P, #Scalasca & CUBE" is now available. A must-watch for anyone starting with #HPC performance analysis and scalability assessment. ▶️ shorturl.at/R6YyX @fzj-jsc.bsky.social @eurohpc-ju.bsky.social

Webinar | POP CoE performance metrics: a beginner’s guide with Score-P, Scalasca & CUBE

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

shorturl.at

🚀 𝗜𝗦𝗖 𝟮𝟬𝟮𝟲 𝗶𝘀 𝘂𝗻𝗱𝗲𝗿𝘄𝗮𝘆 𝗮𝗻𝗱 𝗘𝗣𝗜𝗖𝗨𝗥𝗘 𝗶𝘀 𝗵𝗲𝗿𝗲! Find us at our partners' booths and at the @eurohpc-ju.bsky.social booth (J30). 📅 June 25 · 13:00 | Catch our project overview at the EuroHPC presentation corner! See you in Hamburg! 🇩🇪 #EuroHPC #ISC26 #HPC

📄 What if improving performance wasn’t about rewriting code? New #EPICURE white paper explores a different approach to #GPU utilisation. The work, carried out on Leonardo at @cineca.bsky.social, achieved the same simulation outcomes with up to 4× fewer GPU-hours. 🔗 shorturl.at/hvkVY

🚒 Project Spotlight: Deep Neural Networks for Public Safety Using #MareNostrum 5, this project detect fire and floodwater in real time and at scale. #EPICURE helped solve GPU memory limits and improve #NCCL performance. The result: faster and more stable. 🔗 shorturl.at/TTxy4 @bsc-cns.bsky.social

🎥 𝗥𝗲𝗮𝗱𝘆 𝘁𝗼 𝘁𝗮𝗸𝗲 𝘆𝗼𝘂𝗿 𝗳𝗶𝗿𝘀𝘁 𝘀𝘁𝗲𝗽𝘀 𝗶𝗻 #𝗛𝗣𝗖 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀? The recording of our #webinar “From theory to practice: HPC performance analysis with Score-P, #Scalasca & #CUBE” is now available. ▶️ Watch the full webinar: youtu.be/CJdW2cylNU0 @fzj-jsc.bsky.social @eurohpc-ju.bsky.social

Webinar | From theory to practice: HPC performance analysis with Score-P, Scalasca & CUBE

YouTube video by EPICURE Project

youtu.be

🚨 Still time to register for Webinar: POP CoE Performance Metrics: A Beginner’s Guide with Score-P, Scalasca & CUBE. Learn how to assess parallel efficiency and scalability using POP CoE metrics and practical #HPC performance analysis tools. 🗓️ TOMORROW 🔗 shorturl.at/ddKsw @fzj-jsc.bsky.social

🚀 𝗠𝗲𝗲𝘁 𝗘𝗣𝗜𝗖𝗨𝗥𝗘 𝗮𝘁 𝗜𝗦𝗖 𝗛𝗶𝗴𝗵 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝟮𝟬𝟮𝟲 EPICURE will be featured at the @eurohpc-ju.bsky.social booth (J30) during #ISC26. Join Virvatuli Uusimäki for an overview of the #EPICURE project. 🔗 Know more: shorturl.at/mB607 @csc.fi #EuroHPC #HPC

📈 𝗬𝗼𝘂𝗿 𝗮𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗶𝘀 𝗿𝘂𝗻𝗻𝗶𝗻𝗴. 𝗕𝘂𝘁 𝗶𝘀 𝗶𝘁 𝘀𝗰𝗮𝗹𝗶𝗻𝗴 𝘄𝗲𝗹𝗹? Discover the POP CoE metrics used to evaluate parallel efficiency and scalability, and learn how to interpret them using #Score-P, #Scalasca & #CUBE in the upcoming EPICURE #webinar. 📅 June 16 🔗 shorturl.at/ddKsw @fzj-jsc.bsky.social #HPC

🎥 Our webinar Introduction to #Parallel Performance #Engineering is now available online. Learn the fundamentals of #HPC performance engineering, including #profiling, tracing, scalability and performance analysis workflows. ▶️ Watch the webinar: youtu.be/EO1CPao4Z38 @fzj-jsc.bsky.social

Webinar | Introduction to parallel performance engineering

YouTube video by EPICURE Project

youtu.be

🚨 Last call to join the webinar: From Theory to Practice: #HPC Performance Analysis with Score-P, Scalasca & CUBE. Learn how to profile applications, identify bottlenecks and interpret performance data through practical demonstrations. 🗓️ TOMORROW 🔗 shorturl.at/ddKsw @fz-juelich.de #EuroHPC

🧑‍💻 𝗧𝗵𝗲 𝗿𝗲𝗰𝗼𝗿𝗱𝗶𝗻𝗴 𝗼𝗳 𝗼𝘂𝗿 𝗿𝗲𝗰𝗲𝗻𝘁 𝘄𝗲𝗯𝗶𝗻𝗮𝗿 𝗼𝗻 𝗠𝘂𝗹𝘁𝗶-𝗚𝗣𝗨 𝗣𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴 𝗶𝘀 𝗻𝗼𝘄 𝗮𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲! This is a must-watch to learn more about high-performance #GPU communication software, focusing on #NCCL and #NVSHMEM. ▶️ youtu.be/DLKN9GqucLw @eurohpc-ju.bsky.social

Webinar | Multi GPU Programming in NCCL and NVSHMEM

YouTube video by EPICURE Project

youtu.be

💻 Our new webinar series is underway, and the second webinar is next week! “From theory to practice: #HPC performance analysis with Score-P, Scalasca & CUBE” will introduce practical performance analysis. 📅 June 9, 2026 🕙 10:00 am (CET) 🔗 shorturl.at/ddKsw @fz-juelich.de #EuroHPC

𝗜𝘁’𝘀 𝗯𝗲𝗲𝗻 𝗮 𝗺𝗼𝗻𝘁𝗵, 𝗮𝗻𝗱 𝗼𝘂𝗿 𝗺𝗶𝗻𝗱𝘀 𝗮𝗿𝗲 𝘀𝘁𝗶𝗹𝗹 𝗮𝘁 𝘁𝗵𝗲 𝗘𝗣𝗜𝗖𝗨𝗥𝗘 𝗛𝗮𝗰𝗸𝗮𝘁𝗵𝗼𝗻! 🤩 Researchers, developers and #HPC users together to code optimisation and scalling. We’re already looking forward to the next one! 🚀 🔗 shorturl.at/gk3lD @inesctec.bsky.social @deucalionacnca.bsky.social @hlrs-hpc.bsky.social

INESC TEC hosted European hackathon dedicated to code optimisation for supercomputers

The event, held as part of the European EPICURE project - with INESC TEC as a partner -, brought together programmers, researchers and HPC users from across Europe

shorturl.at

🚨 𝗟𝗮𝘀𝘁 𝗰𝗵𝗮𝗻𝗰𝗲 𝘁𝗼 𝗿𝗲𝗴𝗶𝘀𝘁𝗲𝗿: Introduction to parallel performance engineering The webinar will introduce key concepts in #HPC performance analysis, including the main factors affecting parallel application performance. 📅 TOMORROW | 10:00 am (CET) 🔗 shorturl.at/A4CYg @fz-juelich.de #EuroHPC

⏰ 𝗟𝗮𝘀𝘁 𝗰𝗮𝗹𝗹 𝘁𝗼 𝘄𝗲𝗯𝗶𝗻𝗮𝗿: 𝗠𝘂𝗹𝘁𝗶-𝗚𝗣𝗨 𝗣𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴 𝗶𝗻 𝗡𝗖𝗖𝗟 𝗮𝗻𝗱 𝗡𝗩𝗦𝗛𝗠𝗘𝗠 There’s still time to register for the upcoming webinar with Jeff Hammond (NVIDIA). 🗓️ TOMORROW | 11:00 (CET) 🔗 shorturl.at/MltUh #HPC #AI #NCCL

🎓 New Webinar Series: Profiling for #HPC The first webinar, "Introduction to HPC Performance Analysis", will introduce key concepts in HPC performance analysis. 📅 June 2, 2026 🕙 10:00 am (CET) Individual registration is available for each webinar ➡️ shorturl.at/A4CYg @fz-juelich.de

🚀 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝘆 | 𝗘𝗣𝗜𝗖𝗨𝗥𝗘 𝗣𝗿𝗼𝗷𝗲𝗰𝘁 𝗮𝘁 @inesctec.bsky.social Join the #EPICURE Project and make a difference in application support on #HPC. 👉 Apply here: shorturl.at/YcpPz 👉 shorturl.at/JXHLt @deucalionacnca.bsky.social @eurohpc-ju.bsky.social

🚀 What happens when your model no longer fits on one machine? A new guide by Emmanuel Kieffer (@luxprovide.bsky.social) explains how to deploy large-scale #LLM inference workloads with vLLM across multiple #GPU nodes using tensor and pipeline parallelism. 🔗 Discover more: shorturl.at/BYNjk