Hey everyone, I am building a custom bare-metal operating system and AI inference engine written entirely in C for maximum hardware portability called ANVIL. To prove exactly how lightweight this architecture is, I am currently running the boot sequence natively on a legacy Intel Core 2 Duo T7200 with only 1GB of RAM. Having just wrapped up my AMD ROCm Certified Associate credential this weekend, my next major milestone is porting my execution pipelines directly to native AMD silicon. If anyone in the community has experience with something similar I would love to connect and trade notes on extreme optimization!