Home

Storage

NVIDIA

 NVIDIA AI storage solutions
Turbocharge AI results and NVIDIA GPU access with more data and accelerated global access
Get the IBM Storage for AI ebook
Isometric graphic of two people accessing data on a touchscreen connected to a secure cloud, servers and applications
Accelerate AI and data-intensive workloads

IBM Storage Scale System 6000 is now a certified NVIDIA Cloud Partner (NCP) for HGX H100/H200/B200 systems, delivering high-performance, scalable IO for demanding AI training and inferencing workloads in the cloud.

To maximize AI performance, Storage Scale System 6000 supports NVIDIA GPUDirect Storage, enabling a direct data path between GPU memory and NVMe or NVMe-oF storage. This removes CPU and DRAM bottlenecks, ensuring faster data access for AI and HPC workloads.

Built for AI, high-performance computing, and analytics, IBM Storage Scale provides software-defined file and object storage, while its hardware implementation, Storage Scale System, is optimized for multi-GPU environments powered by NVIDIA HGX™. This seamless integration accelerates AI, complex simulations, and massive datasets.

 Watch how Continental ADAS improves AI data access with IBM Storage and NVIDIA (2:00)
Explore more
Reference architecture
IBM Storage Scale and Scale System 6000 with NVIDIA DGX SuperPOD
Reference architecture
IBM Storage Scale and Scale System 3500 with NVIDIA DGX SuperPOD
Reference architecture
NVIDIA DGX BasePOD with IBM Storage Scale and IBM Storage Scale System 6000
Benefits
Improve efficiency

Build a sustainable, low-energy and high-performance infrastructure, that is flexible for multiple concurrent workloads.

 Read the IBM Redbooks
Accelerate AI

Create an information supply chain for AI and the flexibility to access multiple data sources with a parallel high-performance platform.

 Read the IBM Redbooks
Unleash the Enterprise Core

Create an open hybrid cloud that modernizes infrastructure with global data unification from edge to cloud.

 Read the IBM Redbooks
Storage systems for NVIDIA Simplify your infrastructure by breaking down data silos to lower costs with one data platform and a single source of truth for your AI workloads. IBM Storage Scale

Leverage a scale-out file and object, software-defined storage platform designed for AI, machine learning and high-performance computing workloads.

 Explore IBM Storage Scale IBM Storage Scale System

Take advantage of an all-flash, hybrid elastic compute and storage building block to create high-performing clusters for IBM Storage Scale in a sustainable IT architecture.

 Explore IBM Storage Scale System IBM Fusion

Experience the easiest way to deploy OpenShift applications and harness watson.x's AI capabilities, while seamlessly integrating virtualization and containerization.

 Explore IBM Fusion Storage for AI

Realize faster time to value for data with a unified storage platform that consolidates file, block and object data services and integrates with existing IT investments.

 Explore storage for AI

Meeting the demands of high-speed AI

Continental accelerates AI training time 70% by optimizing its storage infrastructure.  

IBM named as an industry leader for 9 years in a row
Gartner has recognized IBM as leader in distributed file and object storage for many consecutive years. Get the Gartner® Magic Quadrant™ report
Resources Storage Scale is a key part of the IBM Vela AI supercomputer used to train Granite foundation models. IBM Storage Scale developer edition free trial Target Use Cases for Parallel File Systems Extend to Enterprise AI and Analytics Accelerating workloads with IBM Storage Scale & Storage Scale System
Take the next step

Learn how IBM Storage solutions can help you modernize for next-generation AI workloads without compromising.

 Explore storage solutions for AI
More ways to explore Community