Advance Productivity: PowerScale Storage Certified for NVIDIA DGX SuperPOD

Innovate with generative AI built with Dell PowerScale storage and NVIDIA DGX SuperPOD.

In this era of exceptional technological advancements, artificial intelligence (AI) has emerged as a transformative force across industries. The rise of generative AI, in particular, has unlocked tremendous potential, enabling machines to synthesize and create valuable content. Announced at the NVIDIA GTC global AI conference, Dell PowerScale file storage is the world’s first Ethernet-based storage certified for NVIDIA DGX SuperPOD. This solution helps enable businesses to harness the full potential of AI-enabled applications.

Rise of Generative AI

Generative AI has marked a turning point in the technology landscape, enabling machines to create, replicate and learn from patterns in extensive datasets, with minimal human intervention. From healthcare and manufacturing to finance and entertainment, generative AI has the power to transform industries like never before. As a result, the demand for advanced storage solutions capable of handling the sheer volume of data required by AI workloads has skyrocketed.

Certification

Dell has always been at the forefront of innovation, striving to offer cutting-edge solutions that address the evolving needs of businesses. As Dell announced earlier this year, the certification of Dell PowerScale as the first Ethernet-based storage for NVIDIA DGX SuperPOD marks an advancement in storage compatibility. This solution simplifies AI infrastructure, empowering organizations to unlock the true potential of their AI initiatives.

Martin Glynn, Senior Director of Product Management for Dell Technologies, affirms, “The world’s first Ethernet storage certification for NVIDIA DGX SuperPOD with Dell PowerScale combines Dell’s industry-leading storage and NVIDIA’s AI supercomputing systems, empowering organizations to unlock AI’s full potential, drive breakthroughs and achieve the seemingly impossible.”

“To harness the transformative power of generative AI, organizations are seeking robust and scalable infrastructure that can help unlock insights and drive innovation,” said Tony Paikeday, senior director of AI systems at NVIDIA. “The certification of Dell PowerScale as the world’s first Ethernet storage to operate with NVIDIA DGX SuperPOD gives enterprises greater choice and flexibility in their implementations of scalable AI infrastructure.”

PowerScale – Extreme Performance for Next-Generation Workloads

Dell PowerScale has gained trust and recognition for its exceptional scalability, performance and security built on over a decade of experience. Its certification for NVIDIA DGX SuperPOD further strengthens Dell’s storage solution for organizations seeking to harness the power of generative AI. Here’s why PowerScale stands out:

  • Enhanced network access. Technologies like NVIDIA Magnum IO, GPUDirect Storage and NFS over RDMA are natively integrated into NVIDIA ConnectX NICs and NVIDIA Spectrum switches, accelerating network access to storage. These advanced features further minimize data transfer times to and from PowerScale storage, ensuring faster storage throughput for AI training, checkpointing and inferencing tasks.
  • Maximizing performance. Dell PowerScale offers a new Multipath Client Driver that optimizes data throughput and allows organizations to achieve the high performance required by DGX SuperPOD. This innovative feature enables businesses to accelerate AI model training and inference, facilitating seamless integration with the powerful NVIDIA DGX platform.
  • Scalability. PowerScale’s unique design enables effortless scaling by simply adding new nodes, delivering unparalleled agility to businesses. This flexibility ensures that organizations can adapt and grow their storage infrastructure as their AI workloads expand, avoiding any potential bottlenecks for even the most demanding AI use cases.
  • Federal-grade security. PowerScale has been recognized for its exceptional security capabilities, earning the U.S. Department of Defense Approved Product List. Successful completion of this rigorous process underscores PowerScale’s suitability for mission-critical applications and bolsters protection of critical data assets.
  • Efficiency. PowerScale’s innovative architecture is designed to optimize efficiency. Utilizing advanced technologies, it maximizes performance and minimizes power consumption while reducing operational costs and environmental impacts.

To accelerate the deployment of PowerScale with NVIDIA DGX SuperPOD solutions, Dell developed the following reference architecture. The whitepaper illustrates the seamless combination of PowerScale with NVIDIA DGX SuperPOD.

The certification of Dell PowerScale as the world’s first Ethernet-based storage solution certified for NVIDIA DGX SuperPOD represents a significant milestone in the rapidly evolving AI landscape. Dell’s collaboration with NVIDIA is helping to pave the way for organizations to leverage generative AI with scalability and performance. By embracing these leading technologies, businesses can unlock the full potential of AI and transform their operations, propelling themselves into an era of accelerated innovation and remarkable growth.

Richard Greenwalt

About the Author: Richard Greenwalt

Richard is a Product Marketing Manager for Dell’s Unstructured Data Solutions portfolio. In this role, he covers a variety of topics including storage solutions, GenAI, cybersecurity and management software. Richard uses his experience in the storage and server markets to solve IT challenges and simplify data center management.

Richard holds an MBA from the University of Texas at Austin, a Master's Degree in Electrical Engineering from the University of Michigan, Ann Arbor, and a Bachelor's Degree in Electrical Engineering from the University of Texas at Austin.