Cross-vendor byte-identical inference for a 72B LLM (AMD MI300X vs. Nvidia H100)

A new protocol generates byte-identical outputs from frontier-scale language model inference, linking each output to a portable, offline-verifiable signed receipt. This construction features an inference substrate capable of running models with up to 72 billion dense parameters on NVIDIA H100, with AMD's MI300X included. The output hashes match byte-for-byte across configurations, ensuring issuer-independent fabrication soundness at production costs.