NVDA · News · 20261105N

Datacenter Networking Partnership

NVIDIA CORP · 2026-11-05 · Importance 34 · Surprise 42

Supermicro, IBM, and NVIDIA launched an integrated KV-cache storage solution for AI inference workloads. The system combines Supermicro Petascale storage servers, IBM Storage Scale, NVIDIA Dynamo inference software, and NVIDIA Spectrum-X Ethernet switches. The companies reported up to 56-times faster time-to-first-token for long-context prompts, positioning the offering to address infrastructure bottlenecks in agentic and long-context AI workloads.

Key facts

• Supermicro, IBM, and NVIDIA partnered to launch a KV cache storage solution featuring NVIDIA Dynamo inference software and NVIDIA Spectrum-X Ethernet switches. source
• The Supermicro, IBM, and NVIDIA KV cache solution demonstrated up to 56x faster Time-to-First-Token for long-context prompts. source