Allegro AI has joined the NVIDIA DGX-Ready Software program. The combination of the Allegro AI end-to-end ML/DL lifecycle management solution and NVIDIA DGXTM systems, including the newly announced NVIDIA DGX A100, and NVIDIA NGC, GPU-optimized AI software hub, offers both an open-source and enterprise-grade solution for development and deployment of AI workloads.
Allegro AI’s ML-Ops easy-to-use orchestration and scheduling capabilities enable customers to get the most out of their DGX systems, thereby increasing both team productivity and infrastructure utilization.
The Allegro AI enterprise platform is based on the Allegro Trains open-source ML/DL experiment manager and ML-Ops solution. The platform enables data science and data engineering teams to easily develop and scale their AI model training, eliminating the need to address the complexities of specific hardware environments and tedious parameter setup.
Data scientists can focus on innovation while data engineers can deliver at scale by deploying on powerful NVIDIA DGX systems directly from the Allegro Trains web interface. Allegro Trains works directly on bare metal or integrates to leverage an existing Kubernetes or Slurm cluster.
From developer workbench to workflow management to model deployment, the NVIDIA DGX-Ready Software program offers proven solutions that complement NVIDIA DGX system deployment with improved data science productivity, accelerated AI data pipeline and workflow, and improved accessibility and utilization of resources.
“As more organizations use AI at scale, streamlining AI workflows becomes critically important,” said John Barco, senior director of DGX software product management at NVIDIA. “Customers can use the enterprise-grade NVIDIA DGX systems and Allegro Trains to accelerate and scale AI projects on their deployments.”