Eviden, the Atos Group business leading in advanced computing, announces JARVICE AI, a new software solution that complements its full BullSequana AI product and services line. JARVICE AI simplifies the work of data scientists and AI infrastructure administrators by providing a ‘single and secure pane of glass’ orchestration access to geographically dispersed GPU clusters or cloud providers.
JARVICE AI is a comprehensive extension of Eviden’s existing JARVICE™ XE enterprise HPC platform. With new capabilities, this platform is designed for AI and can orchestrate one-click MLOps frameworks on advanced AI infrastructures, whether on-premises, in the cloud, or on federated architectures. The offering boasts bare metal performance and flexibility to run GPU, CPU, IPU or any other emerging AI acceleration resources. JARVICE AI leverages the extensive HyperHub™ application marketplace, and expands it with a catalog of AI-specific applications and frameworks across the Generative AI, model ops and LLM landscapes. Additionally, JARVICE AI can support custom AI applications and frameworks.
With its built-in advanced workload scheduler, JARVICE AI accommodates a diverse set of concurrent AI scheduling needs from a single platform, including large-scale training for generative AI frameworks, low-latency inference tasks, and complex MLOps workflows, delivering high levels of efficiency for cluster utilization and workload mobility.
Also Read: Aetina Introduces AIP-KQ67 for Enhanced Edge AI Computing and Inference
“By combining and simplifying the management and deployment of AI and HPC applications under a single pane of glass across all organizational compute endpoints, JARVICE AI provides a secure and scalable user experience for scientists, researchers, and engineers. With JARVICE AI and HyperHub, organizations can respond to the rapidly evolving needs of AI workloads with confidence and ease – no matter what the future brings,” said Leo Reiter, CTO Nimbix Cloud.
“JARVICE AI excites me as a 6-year user of JARVICE for training large-scale deep learning models. It’s easy to use, unifies MLOps workflows across any infrastructure into a single pane of glass, and introduces new, advanced workload scheduling—it’s quite the compelling offering for the enterprise,” said Tanmay Bakshi, ML & Software Architect, IBM Automation, and Google Developer Expert for Machine Learning at IBM.
Source: GlobeNewsWire