SUSE AI Factory with NVIDIA is a Private Enterprise AI solution that natively embeds NVIDIA AI Enterprise directly into SUSE’s sovereign, hardened infrastructure.
Built on top of the popular LMCache project, Tensormesh Platform enables and manages an enterprise-ready KV caching layer, which is the most effective and efficient way of optimizing inference costs and latency.
Together, they address two of enterprise AI’s toughest challenges: securing data and models to meet compliance requirements, and keeping inference costs predictable at scale, especially in regulated industries and in RAG and agentic workloads.
What the joint solution delivers
Cost reduction
Tensormesh Platform cuts AI inference compute costs by managing your KV cache intelligently across GPU memory, CPU memory and external storage, from prefix and non-prefix caching to prefill/decode disaggregation and cache sharing across GPU nodes.
Ease of deployment and enterprise support
Helm chart–based installation gets you running quickly, and a web UI paired with a command-line interface covers both Day 1 setup and Day 2 support of your AI inference stack.
Multi-vendor support
Built to run on the infrastructure you have: accelerators, inference engines and any storage provider. A single data management layer for AI inference, tightly integrated with SUSE, with no vendor lock-in.
