Atos-owned Eviden has announced a new BullSequana AI product line which includes five AI server models that the company says have been optimized for AI workloads.

Each model offers varying levels of performance, scalability, and flexibility, depending on customers’ specific use cases, budget, and business size.

Such use cases include inferencing AI models at the Edge, fine-tuning existing models in enterprise data centers, or training foundation models with AI supercomputers.

Jean Zay Eviden extension
A prototype of the Jean Zay Eviden extension – Eviden

The flagship product from the server line is the BullSequana AI 1200H, an enterprise-grade supercomputer designed for enterprise use for LLM and complex system modeling. It comes equipped with Nvidia H100s and Nvidia Quantum-2 InfiniBand networking and is cooled by Eviden’s Direct Liquid Cooling technology.

Other products in the server line are the BullSequana AI 800, 600, 200, and 100.

Moving forward, the company plans to integrate Nvidia GB200 Superchips into the server, however, Eviden did add that at an unspecified future date, the 1200H product family will also be able to incorporate providers for AI workloads.

The BullSequana product line also uses business-case-focused AI software, including the company’s own Computer Vision VISuite and Nvidia’s AI Enterprise cloud-native software platform.

France’s HPC agency GENCI and its National Center for Scientific Research (CNRS) are the first customers to purchase the BullSequana AI 1200H. As was announced in March, Eviden is supplying 1,456 Nvidia H100 GPUs to extend the capacity of their Jean Zay supercomputer.

The upgrade will increase the peak computing power from 36.85 petaflops to 125.9 petaflops and will be fully available to users early this summer.

First deployed in 2019, the French supercomputer was named after a French politician imprisoned by the Vichy government during World War II.

“AI is a transforming force capable of revolutionizing businesses across industries and sectors by enhancing productivity and accelerating innovation. However, developing AI applications requires a great deal of expertise, high computing power, and energy to design, train, and run AI models,” said Cédric Bourrasset, head of quantum & AI computing, Eviden, Atos Group.

“This is particularly true for foundation models. With Eviden’s new offering, customers who require reliability, scalability, and versatility to stay ahead of innovation will have maximized acceleration and the right expertise in AI to enable them to make their ideas and innovations a reality,” he added.

Atos, the beleaguered French IT infrastructure and consulting company, has struggled financially for several years. Earlier this week it was revealed that Thales might buy the company’s defense business, which provides Edge devices and other IT infrastructure, as well as software, to military clients.

The French government has also announced plans to purchase key assets from the troubled IT company.