# Silicon

> An open research program exploring how autonomous engineering agents can design, verify, and improve specialized processors, from RTL to physical silicon.

Canonical page: https://aerospace-superintelligence.com/silicon/

Status: founding phase. Milestones are intentions, not achieved results.

## Mission

Make useful processor architecture discoverable, testable, and eventually manufacturable through autonomous engineering.

Research question: can an AI research system propose a hardware architecture, verify its correctness, and demonstrate measurable gains over an honest baseline?

AI-generated engineering ideas are treated as hypotheses, not accomplishments. No performance claim is published without a traceable experiment.

## Research agenda

1. **Domain-specific compute**: small RISC-V extensions, low-power accelerators, and fixed-latency datapaths for real-time systems.
2. **Agent-directed EDA**: structured agent proposals connected to RTL generation, testbenches, synthesis, and design-space exploration.
3. **Measured performance**: area, timing, estimated power, and functional correctness compared with versioned artifacts.

Research cycle: design → verify against independent reference models → measure with reproducible toolchains → keep only evidence-backed improvements.

## Roadmap

- I. Foundation (current focus): reproducible RTL baseline, CI, and formalized design constraints.
- II. Search (planned): automated architecture candidate generation and selection.
- III. Prototype (future): FPGA validation of selected designs.
- IV. Silicon (long-term): assess tape-out options after measured feasibility.

Sister program: [Control discovery](https://aerospace-superintelligence.com/control.md).