The new Dell PowerEdge XL9680L supports up to 72 NVIDIA Blackwell GPUs in a single rack, complete with liquid cooling.
Among all the announcements around artificial intelligence, Dell has unveiled the new PowerEdge XL9860L server. A Storagereview family favorite, the new PowerEdge XL9680L server supports eight NVIDIA Blackwell GPUs in a compact 4U form factor, providing the highest rack-scale density for NVIDIA GPUs in an industry-standard x86 rack. It offers 33% more GPU density per node, 20% more PCIe Gen. 5 slots, and double the North/South network expansion capacity.
The PowerEdge XE9680L features direct liquid cooling (DLC) for enhanced efficiency and greater cooling capacity for CPUs and GPUs. It is also designed for easy serviceability and will be available fully configured with advanced factory integration for rack-scale deployments and on-site installation.
Dell introduced greater density and highly energy-efficient rack-scale solutions for large GPU-accelerated deployments. These solutions include variants such as an air-cooled design supporting 64 GPUs in a single rack and a liquid-cooled format featuring 72 NVIDIA Blackwell GPUs in a single rack.
Accelerating Edge AI Application Deployment
Dell NativeEdge is the first edge orchestration platform to automate the delivery of NVIDIA AI Enterprise software. This platform helps developers and IT operators easily deploy AI applications and solutions at the edge. With new deployment blueprints, businesses in sectors like manufacturing and retail can quickly and accurately analyze their edge data using NVIDIA Metropolis video analytics, NVIDIA Riva speech and translation capabilities, and NVIDIA NIM inference microservices.
Simplifying AI Application Development and Deployment
Dell also unveiled several development and deployment solutions to make it easier for organizations to realize a faster time to value. Those solutions include:
- Dell Generative AI Solution for Digital Assistants: This new solution speeds up the deployment of digital assistants that provide personalized self-service experiences for end users. Implementation Services for Digital Assistants assist organizations in designing, planning, implementing, testing, and scaling the solution.
- Full Stack Deployment Automation: Solutions from the Dell AI Factory with NVIDIA help organizations rapidly establish AI environments for various use cases. Full stack deployment automation, engineered in collaboration with NVIDIA, reduces the time to value by up to 86% compared to do-it-yourself approaches. Combining this with NVIDIA NIM inference microservices further reduces the time from delivery to running inferencing jobs. NIM microservices offer enterprise developers production-ready, optimized inference engines for popular AI models from NVIDIA and its partner ecosystem.
- Dell Accelerator Services for RAG on Precision AI Workstations: These services help shorten the AI development cycle and improve AI application performance. Utilizing RAG on a Dell Precision workstation with NVIDIA AI Workbench, a development platform for experimenting, testing, and prototyping AI and ML projects in a secure environment, Dell provides tailored large language models to expedite AI development.
The expansion of the Dell AI Factory with NVIDIA represents a significant step forward in AI infrastructure and service offerings. By integrating advanced technologies and providing comprehensive support, Dell Technologies and NVIDIA enable organizations to harness AI’s full potential, driving innovation and achieving faster time to value.
Availability
- Dell PowerEdge XE9680L servers and NativeEdge deployment blueprints for NVIDIA will be available in the second half 2024.
- Dell Generative AI Solution for Digital Assistants and Implementation Services for Digital Assistants are available now in North America.
- Dell AI Factory with NVIDIA automated deployment capabilities enabled through Dell professional services will be available in the second half 2024.
- NVIDIA NIM microservices are available today.
- Dell Accelerator Services for RAG on Precision AI Workstations will be available in late May in the United States.
Engage with StorageReview
Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed
