Bull

In 1961, Bull acquired a 20-hectare site in Angers. Inaugurated two years later and described at the time as “the company’s tenth factory, destined to become its first”, the site has continuously evolved ever since.   Discover how the site evolved over six decades, making Anjou the first French region in electronics, while navigating between the production of various successful machines and the complexity of adapting to clients' demands while growing under pressure.   Today, Angers stands as the only supercomputer factory in Europe. Check our blog to discover its history and the challenges it overcame to reinvent itself, as it prepares to open a new production facility this year.

Europe's digital sovereignty debate is gaining momentum, and the conversation must extend beyond the cloud alone. In his latest blog, our CEO shares why strengthening Europe's position requires a broader approach, spanning infrastructure, platforms and strategic technologies across the full digital stack. Read his perspective and discover Bull’s idea on how to accelerate Europe's path towards greater technological sovereignty.

#WeAreBull, and we're proud of what we've achieved together but we're only getting started. As we and our ambitions continue to grow, we're looking for people who want to join us on our journey. Swipe through our open roles – both tech and non-tech Links to job offers are in the comment section Recruitment colleagues: Mélissa Huguette SY SAVANE Donnie Hristova Rémy Grosjean

In this new episode of Bull Voices, meet Michal Šolc! From studying AI in Prague to becoming a senior consultant in AI and data, he takes us through his career path and what a typical day looks like in this fast-moving field. ❔ How did he move from software engineering into AI? How has the field evolved with the rise of LLMs and AI agents? And what does it take to bring AI agents into high-stakes industries like pharma or insurance? Discover how the AI landscape has impacted the AI consultants' missions and the skills that will matter most going forward. Our first episode in English. 🎧 Links in the comment!

Pricing changes faster than it is possible to manage manually. Demand shifts, competitors move, seasons turn - and across thousands of products in multiple markets, spreadsheets stop keeping up. AI Dynamic Pricing addresses that problem directly: it sets the right price in real time, using machine learning to model how demand actually responds to price. AI Dynamic Pricing is purpose-built to move beyond traditional pricing methods. By modelling demand elasticity, seasonality and product relationships, it helps companies optimise every price, every day: improving margins, streamlining pricing workflows and reducing operational risk. Covering the full pricing lifecycle, from data ingestion and modelling to deployment of recommendations into production systems, AI Dynamic Pricing enables organisations to scale AI pricing reliably, with full governance. Swipe through to see the results a major European pharmacy chain achieved 👉

Growing AI adoption relies on massive data flows, driving demand for network infrastructure capable of moving ever-growing volumes of data efficiently, at scale. That's why Bull and Kalray are partnering to develop the networking technologies for AI and HPC infrastructures.   The collaboration will help evolve BXI, Bull's network technology, towards full Ultra Ethernet compatibility - the open standard now adopted across the AI cloud infrastructure - while preserving the ultra-low latency and advanced congestion management that advanced computing depends on.   Built on European IP, the partnership also aims to strengthen technological sovereignty by securing critical value chains and offering an alternative to non-European technologies.   The result: a more sovereign and unique approach in the market to deliver performance, scalability and openness required by tomorrow's AI Gigafactories, data centres and supercomputers, without compromise.   #HPC #SovereignAI #AIInfrastructure #BullAIDestiny

À mesure que l’IA se déploie à grande échelle, les besoins en infrastructures réseau explosent : plus de données à transporter, plus vite, plus efficacement et à moindre coût.   C’est précisément dans cet objectif que Bull et Kalray s’associent : développer les technologies réseau qui équiperont les infrastructures IA et HPC de demain.   Cette collaboration contribuera à faire évoluer BXI, la technologie réseau de Bull, vers une compatibilité totale avec Ultra Ethernet – le standard ouvert désormais largement adopté dans l’écosystème des infrastructures cloud pour l’IA – tout en préservant l’ultra-faible latence et les mécanismes avancés de gestion de la congestion, deux caractéristiques indispensables au calcul intensif.   S’appuyant sur des technologies conçues en Europe, ce partenariat vise également à renforcer la souveraineté technologique en sécurisant les chaînes de valeur critiques et en proposant une alternative à des technologies non européennes.   Résultat : une approche plus souveraine et unique sur le marché, combinant performance, évolutivité et ouverture pour répondre aux besoins des futures AI Gigafactories, des centres de données et des supercalculateurs, sans compromis. #HPC #SovereignAI #AIInfrastructure #BullAIDestiny

A few days on from AMD Advancing AI in San Francisco, and we're still thinking about the conversations. 🤝 Being an Innovator Sponsor of the event meant a week of standing shoulder to shoulder with the people building the next wave of advanced AI infrastructure – during executive keynotes led by CEO Dr. Lisa Su, technical breakout sessions, developer workshops and partner showcases. Dr. Lisa Su's keynote set out a vision of AI infrastructure built around balanced systems, where compute, networking, memory and energy efficiency are designed to work together rather than scaled independently. Helios, AMD's new rack-scale platform, sat at the heart of it. This week, we're proud to announce we're bringing AMD Helios™ rack-scale AI systems to Europe, expanding what's possible for cloud providers, NeoClouds and the next generation of European AI infrastructure. Helios will be deployed at CINES, with GENCI, in preparation for the Alice Recoque exascale project. Thank you to everyone who stopped by the booth, and to the AMD team for bringing this ecosystem together. #AdvancingAI Bruno Lecointe Prashant Sharma Cristina White David Giorgi

A major bank ran an AI pilot that worked flawlessly: speech-to-text and LLMs monitoring call-centre compliance across thousands of agents. ...and when they calculated the cost of scaling their AI solution in the cloud, their "eyes almost fell out."👁️ That's a wall a lot of enterprises are hitting now: the pilot proves the value and then the cloud cost estimate makes scaling almost impossible.  Petr Bednarik, Enterprise AI Global Lead at Bull, joins host Kevin Jackson on the episode of the Scaling Intelligence podcast to dig into how this played out, amongst other topics like national AI factories, hybrid sovereignty, and the full-stack thinking behind them. Link to the full episode in the comments. 👇 #BullAIDestiny #SovereignAI #EnterpriseAI Intersect360 Research

⚡Demand for high-performance computing keeps growing, and so does the question that comes with it: can machines this powerful genuinely be made eco-responsible? The energy footprint of a supercomputer isn't decided by its hardware alone - it's also determined by how the infrastructure is designed, how resources are actually used, and how efficiently the code runs on it. ♻️Each of these aspects has real impact: capturing heat directly at the components – in our case, with an equivalent of a glass of water per compute blade - turns it from waste into a resource that can even warm homes. Measuring in detail where and when energy is consumed exposes waste that would otherwise stay invisible. Finally, optimising a simulation to run four times faster cuts energy consumption with no new hardware involved. So, can supercomputers be eco-responsible? We believe they can, when sustainability is applied as a discipline across the entire lifecycle and across all layers, from components to code. More on this in the first part of a two-part article, linked in the comments. 👇 #BullAIDestiny #HPC #SustainableAI #Exascale