NetApp, NVIDIA gear up deep learning with ONTAP AI architecture to deliver speed and scale that AI demands

NetApp introduced Wednesday NetApp ONTAP AI proven architecture, powered by NVIDIA DGX supercomputers and NetApp AFF A800 cloud-connected all-flash storage to simplify, accelerate, and scale the data pipeline across edge, core and cloud for deep learning deployments and to help customers achieve real business impact with AI.

Although many organizations are adopting new AI platforms, tools, and practices, most don’t have enough control over their distributed data stores to ensure that complete, current, and accessible data is available for their AI projects. Success with AI depends on a business’s approach to data. To be effective in AI use cases and future-proof a business for new AI applications, organizations must achieve visibility into and control of their data, from edge to core to cloud.

By leveraging the NetApp Data Fabric, ONTAP AI enables enterprises to create a seamless data pipeline that spans from the edge to the core to the cloud. This pipeline integrates diverse, dynamic, and distributed data sources, with complete control and protection. With massive processing power and capacity, ONTAP AI removes performance bottlenecks and enables secure, nondisruptive access to data from multiple sources and data formats.

“The combination of NVIDIA DGX and NetApp all-flash arrays meets the infrastructure challenges of today’s AI deployments,” said Jim McHugh, vice president and general manager of Deep Learning Systems at NVIDIA. “NVIDIA’s leadership in AI and GPU computing, combined with NetApp’s innovation in all-flash storage systems, gives customers a faster way to deploy AI with proven performance and a simplified operations model.”

ONTAP AI is a joint NetApp and NVIDIA proven architecture offering that combines the world’s most powerful GPU solutions with the world’s fastest and most cloud-connected flash systems and software.

Features and benefits include easy deployment, getting started faster with AI by eliminating design complexity and guesswork; scaling without limits so that organizations starting out with deep learning can begin with a 1:1 configuration and scale out as their data grows, to a 1:5 configuration and beyond. The high-availability solution is designed with redundant storage and network and server connections.

ONTAP AI testing uses ImageNet data with a NetApp AFF A800 system and NVIDIA DGX-1 servers in a 1:4 storage-to- compute configuration achieved training throughput of 23,000 training images per second (TIPS) and inference throughput of 60,000 TIPS. With this configuration, users can expect to get over 2GBps of sustained throughput (5GBps peak) with well under 1ms of latency while the GPUs operate at over 95 per ent utilization.

A single AFF A800 system supports throughput of 25GBps for sequential reads and 1 million IOPS for small random reads at under 500-microsecond latencies for NAS workloads. These results demonstrate available performance headroom that can support many more DGX-1 servers as requirements increase.

ONTAP AI allows consumers to start small and grow as needed. Add compute, storage, and networking to clustered con gurations without disrupting ongoing operations. Start with a 1:1 storage-to- compute configuration and scale out as data grows to a 1:5 configuration and beyond. NetApp’s rack-scale architecture allows organizations to start with an AFF A220 and grow as needed to scale from hundreds of terabytes to tens of petabytes with all-flash. With NetApp ONTAP FlexGroup, up to 20 petabytes of single namespace can handle more than 400 billion les.

ONTAP AI leverages the NetApp Data Fabric to unify data management across the pipeline with a single platform. It adopts the same tools to securely control and protect data in flight, in use, or at rest and meet compliance requirements.

“Companies of all sizes and across various industries are making significant inroads in using AI to solve real-life problems, especially in the enterprise. However, while very exciting, AI models and workloads are not easy to deploy, and many organizations are struggling,” said Ritu Jyoti, program vice president, IDC. “The NetApp ONTAP AI proven architecture is a powerful solution that will help users accelerate AI-driven business outcomes and overcome deployment obstacles.”

“Developing disruptive AI technology and turning this into breakthrough products and services for our customers is a vital requirement across many markets we work in,” said Monty Barlow, head of artificial intelligence, Cambridge Consultants. “NetApp ONTAP AI powered by NVIDIA DGX supercomputers and NetApp all-flash storage is simplifying and accelerating the data pipeline for deep learning.”

“Organizations everywhere are making significant investments in AI, but they often lack the infrastructure required to optimize their data for AI projects and scale their applications and workflows to create the business impact they expect,” said Octavian Tanase, senior vice president, ONTAP, NetApp. “NetApp’s cloud-connected data solutions and new proven architecture with NVIDIA DGX create a single data environment for AI. This gives customers the control, access, and performance they need to provide the right data at the right time at the right location to their AI applications – all at scale and all integrated, managed, and protected by the NetApp Data Fabric.”


Leave a Reply

WWPI – Covering the best in IT since 1980