Hewlett Packard Enterprise (HPE) has revealed an enhancement in its strategic alliance with NVIDIA, aimed at crafting an enterprise-grade computing platform for generative AI (GenAI). This collaboration has led to the creation of a solution that is tailored and pre-configured for AI tuning and inferencing, enabling businesses of varying sizes to efficiently personalize foundation models with proprietary data and implement production-ready applications across diverse environments, from edge to cloud. This integrated offering by HPE and NVIDIA simplifies the intricate process of developing and operationalizing GenAI infrastructures.
Enterprises embarking on developing and deploying GenAI models for various applications, including conversational search, business process automation, and content generation, necessitate a software and infrastructure ecosystem that is both rapidly deployable and accessible, irrespective of the business’s operational location. This new enterprise-grade computing solution for generative AI is a product of the broadened cooperation between HPE and NVIDIA, aiming to deliver comprehensive, ready-to-use AI solutions. These solutions encompass an amalgamation of HPE’s Machine Learning Development Environment, HPE Ezmeral Software, HPE ProLiant Compute, and HPE Cray Supercomputers, along with NVIDIA’s AI Enterprise software suite and the NVIDIA NeMo framework.
Antonio Neri, HPE’s President and CEO, commented on the collaboration, stating, “Our alliance with NVIDIA places us in an ideal position to offer a fully integrated AI-native solution that significantly streamlines the process of AI development and deployment, with a range of pre-configured offerings.” He emphasised, “The strategic partnership is set to lower the hurdles for enterprises eager to revolutionise their operations with AI.”
Jensen Huang, NVIDIA’s Founder and CEO, remarked, “The surge of the generative AI age is underway, with enterprises vying to redefine their operations. The extended partnership with HPE will empower these enterprises to achieve remarkable efficiency enhancements through AI applications that synthesize business data to fuel precise assistants, informed chatbots, and semantic search capabilities.”
Accelerating AI Deployment with a Tailored AI Tuning and Inferencing Solution
The newly unveiled enterprise computing solution for generative AI represents a cutting-edge AI tuning and inferencing data center solution that serves as an optimal gateway for businesses, large and small, to embark on their AI endeavors with a ready-to-deploy solution. The enterprise computing solution for generative AI permits organizations to employ pre-trained foundation models in conjunction with their private datasets to craft production-level applications, such as AI-driven chatbots. Moreover, the introduction of retrieval-augmented generation (RAG) processes further refines the data’s quality and precision for these applications.
Optimized specifically for AI, the solution boasts a rack-scale architecture that integrates top-tier HPE ProLiant Compute DL380a, pre-fitted with NVIDIA L40S GPUs, NVIDIA BlueField-3 DPUs, and the NVIDIA Spectrum-X Ethernet Networking Platform, designed to meet the demands of hyperscale AI. This architecture has been calibrated to fine-tune a 70 billion-parameter Llama 2 model, consisting of 16 HPE ProLiant DL380a servers and 64 L40S GPUs.
HPE’s AI software suite, featuring the Machine Learning Development Environment with new generative AI studio features, allows rapid prototyping and testing of models, while the HPE Ezmeral Software with upgraded GPU-aware functionalities simplifies deployment and expedites data preparation for AI workloads within a hybrid cloud framework.
NVIDIA’s AI software suite offers the AI Enterprise package, which accelerates the development and deployment of production AI with added layers of security, stability, manageability, and support. This suite includes the NVIDIA NeMo framework, guardrailing toolkits, data curation tools, and pre-trained models, all working in concert to streamline enterprise GenAI processes.
HPE Services: Facilitating the AI Transformation Journey
HPE Services has now expanded its offerings to include a comprehensive range of consulting services, workforce education, and deployment strategies. These new AI services guide customers through the entire process – from GenAI and Large Language Model (LLM) exploration to execution – assisting them in developing the most effective operational models and hybrid cloud data strategies required to construct, implement, and scale transformative solutions. These extensive services are bolstered by the inauguration of new Global Centers of Excellence for AI and Data in countries such as Spain, the United States, Bulgaria, India, and Tunisia.
