This strategic collaboration marks an important step in our ongoing mission to help organizations maximize the value of their data in an AI-driven world. By combining Cloudian’s enterprise-grade object storage capabilities with NVIDIA’s cutting-edge AI and accelerated computing technologies, we’re creating solutions that don’t just store data efficiently but enable faster, more intelligent data processing and analysis.
Looking forward, we expect to be the first provider to receive the NVIDIA Storage Certification for object storage, delivering to customers a validated solution that scales both to exabytes of capacity and to TiBs per second of aggregate bandwidth.
By joining the NVIDIA Partner Network, Cloudian will showcase how its HyperStore platform addresses the distinct requirements of both AI training and inferencing workflows, offering significant advantages over traditional file-based solutions.
Optimized for AI Training Workloads
Cloudian’s HyperStore platform delivers superior capabilities for the demanding requirements of AI model training:
- Complete S3 API compatibility enables data scientists and developers to leverage the existing AWS SDK ecosystem and S3 tools, significantly reducing development time and accelerating time-to-value
- Substantially lower total cost of ownership compared to parallel file solutions, with savings of up to 70% when factoring in acquisition, operation, and scaling costs
- Industry-leading performance with over 30GiB/sec per 1U server with NVIDIA GPUDirect integration, scalable to terabytes per second of bandwidth to match the most demanding training requirements
Purpose-Built for AI Inferencing
For production AI deployments, Cloudian offers advantages that ensure reliable, secure, and scalable inferencing operations:
- Geo-distributed architecture with multi-tier storage (flash and HDD) provides optimal data placement and access across global deployments
- Enterprise-grade security and financial-market certified object lock capabilities ensure data sovereignty and comprehensive lifecycle management
- Support for AIDP GPU acceleration inside the storage system enables inferencing at the storage layer, dramatically reducing data movement
- Non-disruptive, elastic scalability from terabytes to exabytes without application interruption, allowing AI systems to grow seamlessly with business needs
What This Means for Customers
The collaboration will focus on optimizing Cloudian’s HyperStore object storage platform for gen AI workloads, ensuring seamless integration with NVIDIA GPU-accelerated computing environments. For our customers, this means:
- More efficient data pipelines for AI training and inference
- Reduced time-to-insight for data-intensive applications
- Accelerated innovation for organizations deploying AI solutions at any scale
NVIDIA Storage Certification Program for Object
Cloudian will soon be the first object storage solution to earn the prestigious Nvidia Storage Certification for Object Storage. This follows in the footsteps of our earlier leadership announcement, being the first to launch NVIDIA GPUDirect for object storage.
The NVIDIA Storage Certification program, which currently only exists for file storage solutions, validates that enterprise storage systems meet stringent performance and scalability requirements for AI and high-performance computing workloads. This certification is a prerequisite for partners developing agentic AI infrastructure solutions built on the NVIDIA AI Data Platform, a customizable reference design for building a new class of enterprise infrastructure for demanding agentic AI workloads.
With this certification, Cloudian will be at the forefront of AI-ready storage solutions, ensuring our customers have access to storage systems that are rigorously tested and validated to deliver enterprise-grade AI performance. This aligns perfectly with our mission to help enterprises reduce time, cost, and complexity as they build their AI factories and harness enterprise data for reasoning and responding to complex queries.
As part of the NVIDIA Partner Network, Cloudian gains access to NVIDIA’s technical expertise, training resources, and co-marketing opportunities. This will enable us to develop and deliver validated solutions that help customers overcome the challenges of AI deployment and scaling.