DDN Storage Appliance "Doubles Performance" for Nvidia DGX AI Solutions and Speeds Up Analytics and ML in Cloud "by 100%"

DDN (DataDirect Networks, Inc.) announced its next-gen flash and hybrid data platforms for Nvidia DGX POD and DGX SuperPOD AI, analytics and deep learning computing infrastructure.

Powering thousands of Nvidia DGX systems, including Selene and Cambridge-1 DGX SuperPOD systems, the company offers a range of optimized AI data storage solutions for applications such as autonomous vehicles, natural language processing, financial modeling, drug discovery, academic research, and government security.

The A³I AI400X2 system delivers performance of more than 90GB/s and 3 million IO/s to an DGX A100 system. Available with 250 and 500TB all-NVMe usable capacity, and with the ability to scale orders of magnitude more, it is a performing building block for AI infrastructures.

“DDN has been a market leader in AI, analytics and machine learning for many years and our collaboration with Nvidia is leading the industry in performance, efficiency and ease of management at any scale,” said Dr. James Coomer, VP products. “With our next-gen flash and hybrid DDN AI400X2 storage systems, we are effectively doubling performance, improving ease of use and greatly expanding support for all AI users globally.”

DGX systems with the company’s storage solutions have been implemented by IT organizations worldwide. In 2021, the firm delivered more than 2.5EB of AI, analytics and deep learning flash and hybrid storage solutions in the cloud and customers’ data centers. The company expects to achieve growth in its AI business in 2022.

“Nvidia DGX SuperPOD provides enterprises with a proven, turnkey AI infrastructure solution for powering their most transformative work,” said Charlie Boyle, VP, DGX systems, Nvidia Corp. “From compute to networking to storage, every element of a DGX SuperPOD is selected to ensure it provides powerful performance, and DDN storage keeps pace with the needs of the most demanding AI workloads.”

The company is working with Nvidia on next-gen ‘Reference Architecture’ documents that integrate its AI400X2 appliances with DGX A100 systems. Customers will be able to deploy and scale AI turnkey systems using standard DGX POD and DGX SuperPOD configurations. Backed by Nvidia’s leadership in accelerated computing and DDN’s leadership in AI data management at scale, these integrated systems will deliver faste path to AI implementation to customers across market segments and industries.