DDN A³I platform now comes integrated with NVIDIA DGX-1 to simplify AI data center



DataDirect Networks (DDN) announced this week that it is teaming with NVIDIA to transform data centers with storage and compute solutions that are fully integrated and optimized for AI and deep learning (DL) workloads.

Backed by artificial intelligence (AI) expertise in both storage and computing, the DDN A³I platform with NVIDIA DGX-1 AI supercomputers delivers a validated, pre-configured solution that enables high performance at scale, making it faster and easier for every enterprise to gain data-fueled insights through the power of AI and deep learning.

DDN A³I with NVIDIA DGX-1 provides a turnkey solution that delivers end-to-end parallel architecture providing high throughput, low latency and maximum concurrency in data delivery to applications. Easy to deploy, manage, scale and support, the joint solution, available through AI-specialized resellers, delivers immediate workflow enablement and full saturation of GPU resources.

DDN solutions have scalability built-in, with systems delivering more than 200PB of capacity in a single namespace, and more than 1.4TB/s of performance.

The DDN high-performance parallel architecture delivers data to applications and ensures full GPU resource utilization—even for distributed applications running on multiple DGX-1 systems simultaneously. DDN’s client for DGX-1 containers also engages multiple high-speed data paths to inject the performance of NVMe into containerized applications.

With over two decades of experience supporting data-intensive environments, DDN and NVIDIA have global teams of engineers and technical experts to support customers with optimizing every phase of AI and DL projects: initial inception, solution architecture, systems deployment, customer support and future scaling.

DDN A³I with DGX-1 solutions enable simplified ongoing operations with a fully integrated and optimized environment for AI and DL workloads. The reseller partners offering these solutions are trusted advisors that bring together the combined AI and DL expertise of NVIDIA and DDN to provide turnkey systems that are deployed, managed and scaled, along with single point of contact support to ensure user productivity and to give organizations peace of mind.

Built on the NVIDIA DGX-1 AI supercomputer as it’s compute foundation, this powerful solution delivers over one petaFLOPS of DL training performance, leveraging eight NVIDIA Tesla V100 Tensor Core GPUs, configured in a hybrid cube-mesh topology using NVIDIA NVLink for an ultra-high-bandwidth, low-latency inter-GPU communications fabric.

The DGX-1 architecture overcomes the performance bottlenecks of traditional GPU interconnects and offers linearly predictable performance across multiple GPUs. DGX-1 is powered by the NVIDIA DGX Software Stack which is optimized at every layer, including the most popular DL frameworks and the supporting libraries and drivers, enabling unmatched multi-GPU and multi-system training performance.

To meet the requirements of a variety of workloads, DDN A³I with NVIDIA DGX-1 leverages the DDN AI200 and AI7990 storage appliances. The AI200 is an all-NVME, fully-integrated parallel file storage appliance that delivers 20GB/s of throughput and over 350K IOPS to applications, accelerating even the most I/O intensive workloads.

The AI200 is specifically optimized to keep GPU computing resources fully utilized, ensuring maximum efficiency while easily managing tough data operations.

The AI7990 is a hybrid, parallel file storage appliance that integrates both flash and deeply expandable capacity disk in a unified system for simplicity and flexibility. This integration makes it easy to collocate both hot training data and large libraries while maintaining optimal system efficiency. The AI7990 outperforms competing platforms and delivers the economics of capacity disk for growing data library.

“NVIDIA DGX-1 is a full-stack AI solution for enterprise, purpose-built for the unique demands of AI and deep learning,” said Jim McHugh, vice president and general manager, NVIDIA. “DDN A³I with DGX-1 enables organizations to tap into the power of AI with a turnkey solution that saves time lost on system design, integration, installation and troubleshooting, enabling effortless productivity for AI developers and researchers in every enterprise.”

“Artificial intelligence and deep learning applications are creating some of the most challenging workloads in modern computing history, and are straining traditional compute, storage and network resources,” said Paul Bloch, president and co-founder, DDN. “DDN A³I with NVIDIA DGX-1 is an integrated solution that provides unlimited three-dimensional scaling and improved performance as clusters grow, accelerating the end-to-end AI and DL workflow. DDN A³I with DGX-1 is driving faster iteration and, most importantly, speeding business innovation.”

The DDN A³I with NVIDIA DGX-1 solution is now available for purchase through select US-based value-added resellers, including Meadowgate Technologies, Microway and Groupware Technology, as a validated, easy-to-deploy, turnkey offering. Authorized partners in other regions include GDEP Solutions, XENON and E4 Computer Engineering.

 

Leave a Reply

WWPI – Covering the best in IT since 1980