DDN Launches Faster Array HW and KV Cache SW for AI
Blocks & Files, Wednesday, June 24th, 2026
DDN unveiled the A1400X3M array and NVIDIA Dynamo-integrated KV Cache software to accelerate AI inference.
DDN unveiled the A1400X3M storage appliance delivering 160 million IOPS per rack plus NVIDIA Dynamo-integrated KV Cache software to speed AI inference.
The A1400X3M upgrades from Intel Xeon Ice Lake to AMD Genoa CPUs and PCIe Gen 4 to Gen 5, reaching up to 190 GBps read throughput, a 35 percent gain.
Its KV Cache acceleration extends GPU memory into NVMe storage across a 4-tier hierarchy, claiming 55x faster KV cache loading for large-scale inference.
The release adds bare-metal multi-tenancy, KMIP encryption, file pinning, and NAND-accelerated Hot Pools.
General availability is expected by end of Q3 2026.