Researchers have developed F2Asm, a novel system capable of learning exact SASS encoders for NVIDIA data-center GPUs. This tool enables controlled machine-code rewriting by functioning as an open-source assembler for recent NVIDIA architectures, including Rubin SM107. F2Asm utilizes Gaussian elimination over F2 to construct encoders as vector-valued affine maps, supporting architectures like Hopper SM90/SM90a and Blackwell SM100. The system was trained on a substantial dataset of CUBINs from production libraries and CUDA archives, demonstrating exact reassembly in round-trip tests. AI
RANK_REASON Academic paper detailing a new method for learning SASS encoders. [lever_c_demoted from research: ic=1 ai=0.7]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
