Dell partners with NVIDIA for secure, on-premises generative AI

Applications of AI


Dell Technologies and NVIDIA announce joint initiative to enable enterprises to build and use generative AI models on-premises to quickly and securely deliver capabilities such as better customer service, market intelligence and enterprise search .

Called Project Helix, the solution offers a suite of full-stack solutions with technical expertise and pre-built tools based on infrastructure and software from Dell and NVIDIA. Additionally, it includes a complete blueprint for enabling companies to use their own data and deploy responsibly and accurately generative AI more efficiently.

“Project Helix will provide enterprises with purpose-built AI models to more quickly and securely derive value from today’s vast amounts of underutilized data,” said Vice Chairman and Co-Chairman of Dell Technologies. Chief Operating Officer Jeff Clark said. “With highly scalable and efficient infrastructure, companies can create a new wave of generative AI solutions that can reinvent industries.”

“We are at a historic moment at the intersection of amazing advances in generative AI and the demand of enterprises to do more with less,” said Jensen Huang, founder and CEO of NVIDIA. says. “Working with Dell Technologies, we have designed a highly scalable and highly efficient infrastructure that enables enterprises to transform their businesses by securely using their own data to build and operate generative AI applications. .”

Project Helix simplifies the deployment of enterprise-generated AI using a tested combination of optimized hardware and software, all available from Dell. It provides the ability to transform corporate data into smarter, higher-value results while maintaining data privacy. These solutions help companies drive trusted decisions from their data to grow their businesses and rapidly deploy expanding customized AI applications.

Project Helix supports the complete generative AI lifecycle, from infrastructure provisioning, modeling, training, fine-tuning, application development and deployment, to inference deployment and results streamlining. Validated designs help companies quickly build on-premises generative AI infrastructures at scale.

Dell PowerEdge servers such as the PowerEdge XE9680 and PowerEdge R760xa are optimized to deliver generative AI training and AI inference performance. A Dell server combined with his NVIDIA H100 Tensor Core GPU and NVIDIA networking form the infrastructure backbone for these workloads. Customers can combine this infrastructure with resilient and scalable unstructured data storage such as Dell PowerScale and Dell ECS Enterprise Object Storage.

All Dell Validated Designs enable customers to use the enterprise capabilities of Dell server and storage software with observability through Dell CloudIQ software. Project Helix also includes his NVIDIA AI Enterprise software, which provides tools for customers transitioning their AI lifecycle. NVIDIA AI Enterprise includes over 100 frameworks, pre-trained models and development tools, including the NVIDIA NeMo Large Language Model Framework and NeMo Guardrails software for building topical safe and secure generative AI chatbots. It contains.

Project Helix has security and privacy built into its foundational components, including secure component validation. Securing data on-premises reduces inherent risks and helps businesses meet regulatory requirements.

“Companies are eager to explore the opportunities generative AI tools bring to their organizations, but many have no idea how to get started,” said Bob O’Donnell, president and principal analyst at TECHnalogy Research. said Mr. “By bringing together complete hardware and software solutions from trusted brands, Dell Technologies and his NVIDIA are empowering companies to build AI-powered models that can leverage their unique assets to create powerful customized tools. and provide a head start to improvement.”

Dell Validated Designs under the Project Helix initiative will be available through traditional channels and APEX flexible consumption options starting July 2023.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *