Overview
VAST Data Platform for NVIDIA
The World’s First Enterprise NAS Certified for NVIDIA DGX SuperPOD™
Deliver agile, scalable performance for the most challenging machine learning and deep learning workloads. The VAST Data Platform delivers simple and scalable enterprise file services to modern AI clouds.
Achieve limitless levels of scale and performance with the VAST Data Platform, making large scale AI simpler, faster and easier to manage.
Overview
AI at any scale, made super simple.
The era of AI is well underway. As AI becomes a core element and key enabler of every organization's mission the requirements placed on infrastructure and the teams supporting it are changing. This increased reliance on AI places demands on enterprise IT to deliver production-ready AI with an optimal balance of compute, storage and networking to ensure linearly predictable performance that scales in response to business demand.
VAST Data and NVIDIA have joined forces to provide proven, replicable, and validated AI solutions that can benefit every organization. Organizations starting their AI journey can choose from optimized VAST Data configurations for NVIDIA DGX BasePOD. Experienced teams scaling their AI initiatives can leave behind the challenges of complex legacy solutions and experience enterprise-level uptime and simplicity with NVIDIA DGX SuperPOD integrated with VAST Data.
NVIDIA DGX-H100 SuperPOD
As enterprise AI and HPC adoption grows, organizations are seeking full-stack infrastructure that delivers performance and scale to turn their data into insight. Our collaboration with VAST Data enables customers to use the VAST Data Platform with their NVIDIA DGX SuperPOD environments to accelerate modern enterprise and cloud-based AI initiatives.
Charlie Boyle
Vice President of DGX Systems, NVIDIA
Key Benefits
AI performance meets enterprise simplicity.
The NVIDIA DGX SuperPOD with the VAST Data Platform as a certified data store has the key advantage of enterprise NAS simplicity. In contrast to parallel file system-based architectures, the VAST Data Platform not only offers the performance to meet demanding AI workloads but also non-stop operations and unparalleled uptime all on a system that can easily be supported by enterprise IT teams.
NVIDIA DGX SuperPOD Certified Solution
Simplify infrastructure design & scale with validated solutions for all AI workloads
NVIDIA MagnumIO GPUDirect® Storage (GDS) Support
Full read / write support for a direct data path between GPUs and VAST storage
NFS, SMB, and S3 in a single namespace
Enable multi-protocol workflows and end data management challenges
Scale To Exabytes In a Single Namespace
Gain more insights from massive data sets available in real-time
Maximize Predictable Performance For All AI Workloads
Scale to terabytes per second (TB/s) of bandwidth, with millions of IOPS over Ethernet or InfiniBand
Optimize Costs With All-Flash At Archive Economics
Eliminate the need for tiering with compounding efficiencies to reduce the cost of flash without compromising performance
Resources
Innovation begins with understanding
White Paper
GPU Reference Architecture
NVIDIA DGX-A100 + Universal Storage LightSpeed: reference architecture for AI workloads. Includes machine learning and deep learning benchmarking results.
Blog Post
NVIDIA GPUDirect Storage (GDS): The VAST Data Story
Learn about the impact of GPU acceleration in AI, the introduction of GPUDirect Storage, and VAST Data's collaboration with NVIDIA.