NVIDIA GPU Integration Strengthens AI

FREMONT, California, November 18, 2025 - Penguin Solutions, Inc. (Nasdaq: PENG), a leading provider of high-performance computing and artificial intelligence infrastructure solutions, announced support for OriginAI® solution portfolio for NVIDIA's latest graphics processors.

OriginAI provides sustainable top-tier performance for AI architectures of all scales by simplifying deployment. 

At the beginning of 2026, Penguin’s OriginAI solution will support NVIDIA DGX™ B300 and RTX PRO™ 6000 graphics processors for training and inference workloads. The DGX B300, powered by the Blackwell architecture, offers 72 petaFLOPs of training efficiency and 144 petaFLOPs of output performance. The RTX PRO 6000 will be available as a dual-slot PCIe Gen 5 card equipped with 96 GB of GDDR7 memory, enabling cost-effective scaling of AI deployment.

As a NVIDIA DGX-Ready Managed Services Provider, Penguin Solutions simplifies deployment by combining these advanced graphics processors, validated infrastructure architectures, and expert services, accelerating AI adoption and promoting scaled innovation. Both configurations provide pre-validated faster, more reliable deployments and significantly improve time to value with enhanced solution assurance. 

Dara Ambrose, Vice President of Products and Solutions at Penguin Solutions, stated, "Integrating the latest NVIDIA graphics processors into our OriginAI infrastructure solution is a significant advancement in our ability to help organizations operationalize their AI investments." He added, "OriginAI offers high-performance solutions that enable scalable AI workloads in dynamic technological environments."

The OriginAI solution provides a reliable infrastructure for demanding workloads by combining proven architectures, next-generation hardware, advanced cluster management software, and expert-managed services. Supporting scalable architectures starting from 1/4 pod configurations to 90 pods and beyond, it enables deployments from 64 to over 24,000 GPUs. By combining graphics processors from NVIDIA with industry-leading networking and storage options, OriginAI delivers a comprehensive and high-performance solution with faster time to value. OriginAI architectures include leading storage solutions from DDN, Pure Storage, VAST Data, and VDURA.

To ensure optimal performance of the OriginAI solution, Penguin’s factory burn-in testing and integration environment verify AI cluster performance and confirm production readiness before shipment. By combining these validated architectures with the ICE ClusterWare software and expert managed services, Penguin Solutions can deliver consistent top-tier performance for AI infrastructure.

ICE ClusterWare™ software is a core component of the OriginAI solution, enhancing infrastructure performance while simplifying AI cluster management. The latest software release announced this week at SC25 introduces multi-tenancy, enabling operators to manage complex, multi-user environments more efficiently; it also includes features such as proactive problem-solving and identifying silent component regressions.

The general availability of Penguin Solutions’ OriginAI solution, with the latest NVIDIA configurations and storage validation, is scheduled for early 2026.