Software-defined storage company Cloudian has launched a pay-as-you-go edge AI infrastructure.

The Cloudian HyperScale AI Data platform provides operators with access to faster edge deployments. Available via Amazon Web Services Local Zones, the vendor claims it provides “immediate access” to high-performance AI infrastructure without the need to build data centers.

The localized cloud infrastructure is powered by Nvidia Hopper GPUs, providing some 640 GB of GPU memory. AMD’s third-generation EPYC processors also maake up the underlying hardware, interconnected through some 3,200 Gb/s of Elastic Fabric Adapter (EFA) networking.

Cloudian claims its HyperScale configuration provides real-time inferencing with sub-10 millisecond response times, all while processing sensitive data locally.

“Organizations no longer need to choose between performance, data sovereignty, and cost efficiency,” said Neil Stobart, CTO at Cloudian. “By combining Cloudian's high-performance S3-compatible storage with AWS's GPU-based edge infrastructure, we're enabling enterprises to run sophisticated RAG applications within milliseconds of their end users, with zero upfront cost. This dramatically accelerates AI adoption for organizations that previously couldn't justify the infrastructure investment.”

The Cloudian launch comes as the demand for AI compute power grows, but not every business has the need or cost capabilities to build their own data center stack.

AWS has already launched several Local Zones, which are fully managed by the hyperscaler and built for use by specific organizations or communities.

To further augment the desire for edge infrastructure, AWS added S3 storage offerings to its on-premise cloud offerings, providing localized data storage – a move that aligns with increased demands from some jurisdictions for data isolation and sovereignty requirements.

Cloudian’s latest launch looks to further augment the edge infrastructure offerings, providing operators with a faster way to procure compute.

The vendor claims its new service slashes the six to 12 months usually required for implementing traditional edge AI deployments. Instead, it launches in hours, with users making use of AWS’s pay-as-you-go pricing.

Cloudian touts its HyperScale AI Data platform to support public sector agencies, such as providing secure data processing for healthcare providers.

“Enterprise AI success at the edge requires bringing high-performance compute and storage together where data needs to be processed,” said Justin Boitano, Nvidia’s VP for enterprise AI products. “Cloudian's HyperScale AI Data Platform integration with Nvidia accelerated computing on AWS Local Zones enables organizations to build and scale intelligent RAG applications to power AI agents and reasoning workloads close to their data.”