Vibe Nano
A vertically integrated AI infrastructure stack, from silicon through agents.
Infrastructure Stack
Five layers forming a complete AI infrastructure foundation
Agent
AI agents that reason, plan, and execute tasks using the infrastructure layers below.
Framework
Model frameworks, serving systems, and distributed training for AI workloads.
Compiler
MLIR-based compilers and code generation for optimal model execution.
Virt
Virtualization stack from hardware (KVM) through hypervisor (VMM) to containers (Sandbox) for agent isolation.
Silicon
GPU architecture and design principles for understanding hardware acceleration of AI.
Why this stack? Each layer exposes the next bottleneck for AI workloads. By building from silicon up, we understand hardware constraints (memory bandwidth, compute), virtualization overheads (isolation, context switching), compiler optimizations (operator fusion, memory tiling), framework efficiencies (kernel selection, communication patterns), and ultimately agent capabilities (reasoning, tool use).
Project Tracker
Status and progress across all layers
| Category | Project | Description |
|---|---|---|
| Silicon | Nano GPU | SIMT + DSA architecture providing Scalar, Vector, Cube / Memory ISA APIs with Linux kernel drivers. |
| Nano KernelGen | Kernel generation framework for optimizing tensor computations. | |
| Virt | Vibe KVM | Educational implementation of the Linux KVM API for learning Intel VT-x virtualization at the hardware level. |
| Vibe VMM | Minimal Virtual Machine Monitor (VMM) in C with clean hypervisor abstraction for cross-platform VM execution. | |
| Nano Sandbox | Educational OCI-compatible container runtime in C with pure-container and Firecracker-backed execution modes. | |
| Nano K8S Tutorial | Hands-on tutorial for building a minimal Kubernetes cluster from scratch using containerd and the K8s control plane APIs. | |
| Framework | Nano Train | Learning-first distributed LLM training framework built around Megatron-style parallelism, covering TP, PP, EP, DP, ZeRO-1/2, and mixed precision on a DeepSeek-style model stack. |
| Nano Torch | Deep learning framework with PyTorch-like API. | |
| Nano Serving | Model serving and inference optimization system. | |
| Nano Agentic RL | An agentic RL post-training framework. | |
| Agent | Nano Coder | Minimal terminal-based coding agent with tool use, streaming answers, MCP support, and local subagents for code execution and development workflows. |
| Nano AgentEvolve | A Self-Evolving Agentic Framework. |