nx_sovgpu_research_fetch.nx

nx_sovgpu_research_fetch.nx -- bank the SOVEREIGN GPU STACK landscape (operator 2026-07-04: "stop using even 3rd-party AIs, nishi ecosystem ONLY, build ALL of it ... get BEYOND webgl2/webgpu, there are lots of open source things that can get us to state of the art" + "T3 sovereign GPU on our own silicon = crack it, be FIRST"). WebGL2/WebGPU are just APIs; the sovereign path is BUILDING the GPU implementation. Two proven open routes: (A) SOFTWARE GPU -- implement the full modern graphics API in software (Mesa llvmpipe / lavapipe / SwiftShader = proof a CPU can BE a GPU), sovereign + runs on our own CPU/silicon TODAY; (B) OPEN GPU HARDWARE -- open GPU designs (Vortex RISC-V GPGPU / Nyuzi / MIAOW / Libre-SOC) for our own FPGA. Plus the API/IR layer (Vulkan/SPIR-V/DRM/Gallium) we'd reimplement, and the parallel model (SIMD/SIMT/shader). Sovereign TLS; idempotent. expect_exit:0 ORIGINAL