The Latest From an Enduring Collaboration
Being one of the world’s largest providers of modern IT infrastructure, Dell Technologies strives to make it easy to design and deploy your valuable IT and data science resources for successful AI initiatives. We do so by offering simplicity and choice in how you build and scale an optimal IT infrastructure. NVIDIA, the world’s leader in hardware and software solutions for AI, has worked with Dell to create an enterprise platform that supports the widest range of AI and other data science workloads: the Dell and NVIDIA Enterprise Architecture for AI at Any Scale.
Over the last several years, the two companies have established a strong relationship as we work to help organizations accelerate their AI initiatives. Our collaboration is built on the philosophy of allowing flexibility and informed choice across extensive portfolios of products and solutions. Together, our technologies provide the foundation for successful AI solutions that support the concurrency, performance and capacity requirements of diverse datasets.
As tested, the new enterprise architecture provides infrastructure that can start small and scale concurrently with data growth. It keeps compute, GPU and storage resources performing optimally while giving seamless, multi-protocol access to larger datasets.
Simplify the Journey to Enterprise AI in Production
This powerful combination of massively scalable storage, hardware-accelerated compute and high-speed network with self-service access for your global teams makes it an optimal foundation for an AI Center of Excellence. The powerful and diverse NVIDIA AI Enterprise software stack, along with Domino Data Labs MLOps platform, simplifies the AI journey for any workload. It enables different groups and departments to run simultaneous workloads in the same infrastructure and can support AI initiatives that grow over time.
Together, Dell and NVIDIA help you move AI to production faster and simplify support by reducing the “unknowns” associated with large-scale technology implementations.
Incorporate a Choice of Best-of-breed Hardware and Software
The technology stack includes a mixture of standard hardware and software, but can flexibly incorporate other options such as alternate hardware, as well as a customized software stack to suit the workload requirements. There are four key elements:
Flexible, scalable multi-protocol storage – helps you access and profit from file and block data with storage that can start small, and then scale and grow:
-
- Dell PowerScale — the world’s most flexible1 and secure2 scale-out NAS solution, tailored for unstructured data, supporting up to up to 30PB per cluster and up to 100GbE connectivity
- Dell PowerStore – intelligent, highly adaptable, high-performance storage tailored for structured data workloads
Hardware-accelerated compute – enables demanding AI workloads to be completed much more quickly than CPU-only infrastructure, and with lower TCO and power consumption:
-
- NVIDIA DGX – DGX A100 systems, for training of AI models
- Dell PowerEdge – NVIDIA-Certified PowerEdge servers with NVIDIA GPUs providing inferencing compute and for application hosting
High-performance networking – for the massive bandwidth required to support today’s most demanding, large-scale AI initiatives:
-
- NVIDIA ConnectX SmartNICs – for high-speed Ethernet and InfiniBand networking
- NVIDIA Spectrum Ethernet and NVIDIA Quantum-2 InfiniBand switches – for high-speed connectivity from storage to compute, as well as between compute nodes
Software platform – fully supported, enterprise-grade software that provides all the capabilities necessary to support full-production AI:
-
- NVIDIA Bright Cluster Manager – for fast deployment and end-to-end management of AI server clusters
- NVIDIA AI Enterprise – for full lifecycle development and deployment of AI
- Domino Data MLOps platform – for unified collaboration
Take the Next Steps
Learn more about how this enterprise-ready AI architecture can help drive your enterprise AI initiatives to production faster, at any scale. See our new Reference Architecture White Paper for clear and complete deployment guidance, as well as the Solution Overview for a summary of business value and features. We also invite you to chat with our AI experts and see our technologies live at SC22, the International Conference for High Performance Computing, Networking, Storage and Analysis. The event continues through Nov. 18 in Dallas, Texas.
1 Based on Dell analysis of publicly available information sources measured across seven attributes, August 2021.
2 Based on Dell analysis comparing cybersecurity software capabilities offered for Dell PowerScale versus competitive products, September 2022.