AI Everything

What Bull is bringing

• Bull at AI Everything

  Bull brings a connected AI portfolio to AI Everything, covering sovereign AI software, rack scale training infrastructure, on premises generative AI and liquid cooled supercomputing. Together, these offerings support organisations that need to run AI workloads at scale while keeping full control of their data, models and infrastructure.

• Explore the sovereign AI platform

  The BullSequana AI platform is a modular software stack built on open-source components for running production AI workloads, from LLM serving and RAG to fine tuning and classical machine learning. It is model agnostic, runs on BullSequana servers or existing NVIDIA GPU infrastructure, and is available in Core and Pro tiers.

• Examine rack scale LLM infrastructure

  The Bull AI SuperCluster is an integrated infrastructure based on the NVIDIA GB200 NVL72, designed for training and inference of large language models. Each rack connects 36 Grace CPUs and 72 Blackwell GPUs with hybrid air and direct liquid cooling, and Bull provides installation, configuration and multilevel support.

• Review on premises generative AI

  GenAI in a Box is an on premises appliance that brings together BullSequana AI servers, the BullSequana AI platform and curated open source or commercial models. Every system is installed and tested at the Bull factory in Angers, France, and includes more than 20 prebuilt accelerators such as chatbots and document assistants.

• Discuss high performance AI architectures

  The Bull high performance AI portfolio combines the liquid cooled BullSequana XH3500 supercomputer with BullSequana AI midrange GPU servers for large language models, AI inference and complex numerical simulation. Bull manages delivery, installation and configuration, then provides maintenance and technical support throughout the lifetime of each system.