Skip to content

IN DEVELOPMENT · NO CHIP HAS BEEN TAPED OUT

Why the design work is being set up to run on Czar Forge

The program is being set up so that its AI Employees work on Element 31's own repositories, on a Czar Forge. The intent is that the chip's design stays on hardware Element 31 controls.

· 4 min read

IN SHORT

BramHA's design work is being set up to run on Czar Forge, which is Czar configured for technology and AI teams. The intent is that the AI Employees work on Element 31's own repositories, on Element 31's own appliance, so the design stays on hardware Element 31 controls. This is a statement of intent, not a completed fact, and no chip exists.

Before it is fabricated, a chip is a collection of files: a specification, descriptions of logic, tests, layouts, firmware, and the plan that ties them together. Those files are the design. Where they are kept, and what works on them, is a question of custody.

BramHA is being set up so that this work runs on Czar Forge. The wording is deliberate. The program is in development, no chip exists, and the arrangement described here is intended, not complete.

What Czar Forge is

Czar Forge is Czar, Element 31's sovereign AI appliance, configured for technology and AI teams. On a Forge, agents work on the team's own repositories, on the team's own appliance.

Its boundary is the one Element 31 keeps for all team work: "Agents on a Czar prepare and propose; a named person approves."

Why a chip program fits

Most of a chip's design is text under version control. The logic is written in a hardware description language. The tests are code. The firmware is code. The scripts that drive the design tools are code. The layout itself is produced by running tools over those files.

So the AI Employees on the program, in their roles of architecture, logic design, verification, physical design, firmware, and program, are intended to do what agents on a Forge do: read a repository, propose changes to it, run checks, and hand the result to a person. AI Employees design the chip describes each role.

In that sense the chip program is a technology team's work, and the program plans to treat it as one.

Why custody matters for a design

There are two reasons.

The first is the design itself. Until a part is fabricated, those files are everything a chip program has produced. BramHA is designed and owned by Element 31, and keeping the work on hardware Element 31 controls keeps the design in Element 31's hands.

The second is the kind of chip. The first design is intended to include a hardware root of trust, the part that would let a machine prove what it is and what it is running before a site lets it on the network. For a part whose purpose is to be trusted, it matters where its design was kept and who could reach it while it was being written.

Neither reason is a claim about the result. They explain why the program is being arranged this way.

What stays the same

Running on Forge does not change who decides. AI Employees propose. A named engineer signs the specification, the design freeze, and the tape-out. No AI Employee approves a gate, and a named engineer signs every gate explains why that matters for a part that cannot be patched after fabrication.

And it does not change what the silicon is for. It is being designed for the machine's side of the link: control, motion, sensing, and identity, beginning with a small real-time controller with a root of trust.

What this article does not say

It does not say the work is running this way now. The program is being set up to.

It does not say which design tools run where, how the repositories are organized, or how much of the design flow has been exercised. None of that has been published.

It does not say the arrangement makes the design better. Where the work runs is a matter of custody, not quality. The quality of a design rests on verification and on the engineer who signs each gate.

And it is not a claim about any other team's use of Forge. It concerns one program, Element 31's own, described as intent.

What in development means explains how to read statements of this kind across the site. The short version applies here as everywhere: there is a plan, there is no chip, and the site says only what the program intends.

Questions this raises

Is the chip being designed on Czar Forge today?
The program is being set up to run that way. That is intent, not a completed fact. No chip has been taped out and no chip exists.
Why does it matter where the design is kept?
Before fabrication, the design is everything the program has produced, and the first design is intended to include a hardware root of trust. For a part whose purpose is to be trusted, it matters where its design was kept while it was being written.
Does running on Forge change who approves the design?
No. The boundary is the same: agents prepare and propose, and a named person approves. On the chip program, a named engineer signs the specification, the design freeze, and the tape-out.

KEEP READING