Digital simulation is evolving from a walled garden of on-premises, high-priced business models toward more flexibility and simplicity in the way computing resources are accessed by end-users. Exascale is imminent, but its promise cannot be just about Exaflopic performance. While the Top500 listing remains a good reference in the HPC world, the Green500 listing has been more closely watched in recent years.
The Future of Supercomputing is Hybrid
We believe that the future of supercomputing is hybrid platforms, combining cutting-edge processing units, whether CPUs, GPUs, IPUs, FPGAs, or QPUs, with the support of our partners. Earlier this year, we revealed our BullSequana XH3000, delivering unparalleled computing power at any scale, up to an exascale-class system for a future-proof platform.
Hybridization is also about leveraging the power of Artificial Intelligence (AI). The introduction of AI has opened a new paradigm of applications in various scientific and industrial domains. AI-augmented solutions are the backbone of making advanced computing faster and more efficient, from hardware to software. In our HPC Software Suites, we have the Smart Data Management Suite. This Suite is a set of tools that monitors, stores and analysis all IO-related events, using machine learning to boost overall application performance by up to 30%, requiring zero application change and is one of the mechanisms enabling us to achieve unprecedented performance with global system efficiency.
Our ThinkAI solution, putting the power of AI simulation at the fingertips of end-users, has continuously gained traction since its launch in 2021. We are proud to have delivered BerzeLiUs, a powerful resource to advance AI research and boost collaboration between academia and the Swedish industry.
Hybridization is also about on-premises and cloud resources and being agile when needed. Creating accessible, simple, and flexible environments that empower users to run their workloads where ever and however they want is the cornerstone of our HPCaaS offering. Through our acquisition of Nimbix, Atos is uniquely positioned to deliver advanced as-a-service capabilities to our customers; with exciting new consumption models being announced during ISC 2022.
Hybridization is also about the acceleration of quantum computing and delivering the quantum-accelerated hybrid HPC. Atos is at the forefront with our Quantum Program initiated 6 years ago. With Quantum Learning Machine (QLM), and myQLM, the essential components of the Atos Quantum Program, we help people learn and experiment with quantum computers and develop applications and algorithms without the need to wait for quantum machines to be physically available. We are proud to see our QLM has been adopted by numerous customers worldwide, such as Total in France, C-DAC in India, and Hartree in the UK. With the newly revealed BullSequana XH3000 and our participation in numerous European initiatives, such as the European HPC-QS project and the Q-EXA project, Atos is targeting to deliver the quantum accelerated HPC as early as possible. With the Open Sequana program, Atos allows for 3rd party technology ecosystem participation and delivers customers a future-proof platform ready for quantum accelerators.
Leader in Tackling Scientific Challenges with Research Institutions in Europe
I want to stress why Atos stands for #1 in HPC in Europe. It is not only because we deliver the most systems to European HPC centers in terms of PetaFlops; we have been working with many European research and scientific institutions to empower many crucial data productions and simulations daily. In addition to the Atos Life Sciences Centre of Excellence and Center of Excellence in Weather & Climate Modelling, we have worked with major European institutions, such as ChEESE, to help researchers translate data into actionable insight. Ultimately, our systems are just the technical means; our objective is to deliver life-changing innovations to every citizen through industrial and scientific breakthroughs.
SVP, Head of HPC, AI & Quantum at Atos