Decide. Compress. Move.
SpinMem optimizes placement. SpinCache compresses and decodes. KX MemoryForge adds its own multi-tier cache runtime. Their shared vocabulary is not proof of a completed integration.
AI INFRASTRUCTURE / FOUR PRODUCTSThe full offering
Beyond
the solve.
Persistent state. Multi-tier cache. Compute-fabric planning. Agent workloads. Four operating layers around AI infrastructure, separate from the Forge15 tool family.
Discuss your infrastructureAvailable nowAll four infrastructure products are available with CLI, MCP and headless access.
View interfacesGPU-NATIVE PERSISTENT STATE
StateForge
Available nowCLIMCPHeadless
A GPU-native persistent-state engine connecting supported live-data, computation, indexing, transaction and persistence workflows.
MULTI-TIER KV-CACHE RUNTIME
KX MemoryForge
Available nowCLIMCPHeadless
A multi-tier LLM KV-cache runtime for compressing, offloading and restoring cache in a supported memory and storage environment.
AI COMPUTE-FABRIC PLANNING
GridForge
Available nowCLIMCPHeadless
AI compute-fabric planning software that maps workloads against power, cooling, memory, network and silicon resources.
AGENTIC WORKLOAD OPERATIONS
AgentForge
Available nowCLIMCPHeadless
An operating layer for generating, scheduling, validating, storing and replaying agentic-AI workloads across CPU/GPU infrastructure.
RELATED, NOT INTERCHANGEABLESeparate responsibilities.
Separate responsibilities.
Considered integrations.
These are not four more optimization solvers. Each has its own deployment role, runtime requirements and qualification scope.
Plan the fabric.
Schedule the work.
GridForge models power, cooling, memory, network and silicon. SpinSovereign handles policy-aware GPU scheduling. An integrated handoff needs its own acceptance plan.
Run the work.
Keep its state.
AgentForge operates workloads and replay. StateForge addresses live and persistent GPU-owned data. Permission, recovery and replay semantics stay explicit.
COMMERCIAL SCOPELicense a layer.
License a layer.
Or integrate a workflow.
New infrastructure products are quoted by proposal. Select the runtime, deployment environment, integration and support you actually need.
WHOLE-SYSTEM OPTIMIZATIONThe connection
The connection
can change the result.
Joint configuration, decision-sensitive computation, checked memory, useful future options, exact reformulation and evidence-led improvement. Explore the role of every product and agree a matched system evaluation.
Customer-specific integration opportunities—not added software entitlements or universal performance claims.