---
title: AI infrastructure | SpinForge
url: https://spin-forge.com/infrastructure
description: Persistent state. Multi-tier cache. Compute-fabric planning. Agent workloads. Four operating layers around AI infrastructure, separate from the Forge15 tool…
---

AI INFRASTRUCTURE / FOUR PRODUCTS [The full offering](https://spin-forge.com/platform)

# 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 infrastructure](https://spin-forge.com/contact?product=infrastructure-products)

Available now All four infrastructure products are available with CLI, MCP and headless access.

[View interfaces](https://spin-forge.com/developers?section=access-directory)

[01

GPU-NATIVE PERSISTENT STATE

## StateForge

Available now CLI MCP Headless

A GPU-native persistent-state engine connecting supported live-data, computation, indexing, transaction and persistence workflows.](https://spin-forge.com/stateforge) [02

MULTI-TIER KV-CACHE RUNTIME

## KX MemoryForge

Available now CLI MCP Headless

A multi-tier LLM KV-cache runtime for compressing, offloading and restoring cache in a supported memory and storage environment.](https://spin-forge.com/kx-memoryforge) [03

AI COMPUTE-FABRIC PLANNING

## GridForge

Available now CLI MCP Headless

AI compute-fabric planning software that maps workloads against power, cooling, memory, network and silicon resources.](https://spin-forge.com/gridforge) [04

AGENTIC WORKLOAD OPERATIONS

## AgentForge

Available now CLI MCP Headless

An operating layer for generating, scheduling, validating, storing and replaying agentic-AI workloads across CPU/GPU infrastructure.](https://spin-forge.com/agentforge)

RELATED, NOT INTERCHANGEABLE

## Separate responsibilities.
Considered integrations.

These are not four more optimization solvers. Each has its own deployment role, runtime requirements and qualification scope.

MEMORY

### 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.

RESOURCES

### 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.

AGENT WORK

### 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 SCOPE

## 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.

[Product licensing](https://spin-forge.com/licensing?tab=forge) [Integration & deployment](https://spin-forge.com/integration) [Embedding & OEM](https://spin-forge.com/partners)

WHOLE-SYSTEM OPTIMIZATION

## 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.

[Explore all six connections](https://spin-forge.com/system-synergies) [64-entry interaction map](https://spin-forge.com/system-synergies/product-map)

Customer-specific integration opportunities—not added software entitlements or universal performance claims.