en.Wedoany.com Reported - DDN has significantly expanded its AI and high-performance computing (HPC) data platform portfolio at ISC 2026, introducing new appliances, distributed cache acceleration technology, and multiple security and efficiency enhancements to address key bottlenecks as enterprises transition from AI pilots to production-grade operations.

At the core of this launch is the new DDN AI400X3M appliance, the latest version of its EXAScaler platform. Designed for the most demanding AI and HPC environments, this appliance delivers up to 35% higher read throughput compared to the previous generation, achieving a maximum throughput performance of 190 GB/s and providing up to 30 PB of storage density within a single rack. The AI400X3M supports hybrid disk configurations to optimize cost-effectiveness and scalability, and is expected to be generally available by the end of the third quarter of 2026.

DDN has also officially released a distributed KV cache acceleration architecture integrated with NVIDIA Dynamo, now available on the DDN Infinia and EXAScaler platforms. This technology accelerates large-scale AI inference by eliminating memory bottlenecks and allowing model context to be retrieved directly from DDN's data platform. For large-scale inference workloads, KV cache loading performance improves by up to 55x, while optimizing AI factory return on investment by reducing per-token costs and improving GPU utilization.

In the cloud AI infrastructure space, DDN highlighted its Managed Lustre innovations in collaboration with Google Cloud Next, as well as new deployments for Salesforce. With Google Cloud Managed Lustre powered by DDN EXAScaler, Salesforce achieved 1.5x faster model training, a 75% reduction in I/O latency, and a 42% decrease in training costs.

DDN also introduced several platform enhancements, including bare-metal multi-tenancy, KMIP-based encryption and key management, integrated VictoriaLogs for operational visibility, intelligent file locking, and a NAND-accelerated hot pool for tiering data from all-flash drives to cost-effective HDDs. These updates aim to improve workload isolation, governance, and infrastructure efficiency.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com