Dell Technologies and Red Hat today announced a significant advancement in the Dell APEX Cloud Platform for Red Hat OpenShift, designed to streamline and accelerate the deployment of AI solutions. This enhanced platform delivers a fully integrated, automated infrastructure optimised for Red Hat OpenShift AI, setting a new standard in how organisations handle AI complexities.
In a rapidly evolving AI landscape, the Dell APEX Cloud Platform now supports an expanded range of GPUs, including the new NVIDIA L40S, to cater to the most demanding AI applications. The platform also introduces support for both Dell PowerFlex and Dell ObjectScale storage solutions, ensuring robust, scalable, and cost-effective management of massive datasets and AI workloads.
Key Enhancements Include:
- NVIDIA L40S GPU Integration: The Dell APEX Cloud Platform now supports the NVIDIA L40S GPU, providing advanced computational power essential for handling the most demanding AI applications. This addition offers increased flexibility, allowing organisations to customise their infrastructure according to their specific AI needs, whether for high-performance computing or complex AI models.
- Dell ObjectScale Storage: In addition to Dell PowerFlex storage, the platform now includes Dell ObjectScale storage, enhancing scalability and cost-effectiveness for managing vast amounts of data. ObjectScale is critical for efficiently handling AI workloads that involve large language models and extensive datasets, ensuring reliable and scalable data management.
- Hosted Control Planes for Red Hat OpenShift: The APEX Cloud Platform now supports hosted control planes, which streamline cluster management and reduce associated costs. This feature optimizes deployment times, separates management and workload concerns, and allows organisations to concentrate on their applications rather than infrastructure management.
- Updated Design for Digital Assistants: The Validated Design for deploying digital assistants has been upgraded from a 7B parameter model to a more robust 13B model. This design uses the retrieval-augmented generation (RAG) framework to augment large language models (LLMs) with company-specific data, enabling rapid adaptation and training with relevant information.
- NVIDIA Riva Integration: The new solution incorporates NVIDIA Riva, a microservice designed for building GPU-accelerated speech AI applications. This integration facilitates automated speech recognition (ASR) and text-to-speech (TTS) capabilities, streamlining the deployment of natural language processing (NLP) solutions and opening new possibilities for AI-driven communication tools.
The updated validated design now includes a 13B parameter model for digital assistants utilising a large language model (LLM) with retrieval-augmented generation (RAG) framework, enabling businesses to rapidly adapt LLMs with domain-specific data. The new AI solution also features automated speech recognition (ASR) and text-to-speech (TTS) capabilities, powered by NVIDIA Riva.
Speaking on Dell’s collaboration with Red Hat, Caitlin Gordon, Vice President, Multicloud Product Management at Dell Technologies said: “In the rapidly evolving landscape of artificial intelligence (AI), seizing opportunities amidst complexity is paramount. Collaboratively engineered with Red Hat, the Dell APEX Cloud Platform for Red Hat OpenShift offers a streamlined and automated turnkey solution that transforms how organizations run Red Hat OpenShift on-premises.”
Stefanie Chiras, senior vice president, Partner Ecosystem Success, Red Hat said, “The rise of AI has led customers to seek out hybrid cloud infrastructure that accelerates AI application development and delivers faster time to value. Innovation is at the heart of our continued collaboration with Dell and the updates announced today showcase why Dell APEX Cloud Platform for Red Hat OpenShift can empower organisations with a more consistent, reliable integrated, and automated platform for running Red Hat OpenShift AI on-premises.”