DDN added new capabilities to the Lustre platform it manages with Google Cloud, including means to share key-value (KV) cache to boost AI inference workloads.

Unveiled at Google’s annual Next event, the pair showcased using Managed Lustre as a shared cache layer across inference clusters rather than keeping KV-cache in each server's local memory. As a result, DDN and its hyperscaler helper claim total inference throughput was improved by as much as 75%.

The parallel file system made its debut in 2025, and lets users deploy and scale Lustre environments for high-performance computing (HPC) and AI workloads without the need to manage the underlying servers.

While customers have been using it to support AI model training and fine tuning, the Next updates focus on boosting inference workloads, as businesses actually begin to deploy systems.

The headline update from the Google event is the use of Managed Lustre as a shared KV cache for inference. Using Lustre’s low latency and high aggregate throughput, the pair contend platform users can “avoid redundant computation and scale inference across clusters with virtually unlimited shared cache capacity” – a potential boon considering memory crunch-conscious users are looking to avoid spending more amid industry-wide price hikes.

DDN and Google Cloud also claim the KV cache update reduces the mean time to first token by more than 40% when compared to just KV Cache in host memory alone.

The pair also flagged improvements to how the platform handles hot and cold data tiers. The new single, dynamic tier offers high performance for hot data (information frequently accessed or modified during workloads) to boost token economics while reducing SKU sprawl.

In the almost 12 months since the platform was made generally available, the pair claim to have witnessed “rapid customer adoption.”

“We are delivering one of the fastest-growing, highest-performance managed Lustre services in the industry, purpose-built for the realities of modern AI at scale,” DDN CEO Alex Bouzari noted. “This announcement reinforces DDN’s leadership in AI data platforms and our shared commitment to helping customers innovate faster, at lower cost, and with greater confidence.”