Metraipolis › archive › dc_47ce4c8cdae8
Learning Exact NVIDIA SASS Encoders with $\mathbb{F}_2$ Linear Algebra — arXiv:2608.20532
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.20532
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T10:31:01.310Z
- SHA-256 of the archived text
575552faae50d106f99055460bfa83eed57dc7e1f0f9791683c03f10b46fc51a- Stable URL of this page
https://metraipolis.com/documento/dc_47ce4c8cdae8
How to cite
Learning Exact NVIDIA SASS Encoders with $\mathbb{F}_2$ Linear Algebra — arXiv:2608.20532 — arXiv (en), https://arxiv.org/abs/2608.20532 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_47ce4c8cdae8 (Metraipolis archive dc_47ce4c8cdae8, sha256 575552faae50d106f99055460bfa83eed57dc7e1f0f9791683c03f10b46fc51a)
BibTeX
@misc{dc47ce4c8cdae8,
title = {Learning Exact NVIDIA SASS Encoders with $mathbbF_2$ Linear Algebra — arXiv:2608.20532},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.20532},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_47ce4c8cdae8; sha256 575552faae50d106f99055460bfa83eed57dc7e1f0f9791683c03f10b46fc51a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.20532. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
NVIDIA provides a SASS disassembler but no public SASS assembler for recent data-center GPUs, limiting controlled machine-code rewriting. We present F2Asm, which learns exact 128-bit SASS encoders from paired disassembly and original CUBIN instruction words. To our knowledge, F2Asm is the first system to learn SASS instruction encoders as vector-valued affine maps over $\mathbb{F}_2$ and the first open-source NVIDIA SASS assembler to support Rubin SM107. F2Asm uses Gaussian elimination over $\mathbb{F}_2$ to incrementally build a compact basis, detect inconsistencies, and reject inputs outside the learned span. F2Asm separates target-specific control bits, relocation rules, and CUBIN metadata from its learning algorithm. We train encoders for Hopper SM90/SM90a, Blackwell SM100, and Rubin SM107 using 3,225 CUBINs from pinned NVIDIA and third-party production libraries, CUDA 13.3 packages, and CUDA 13.4 Developer Preview archives. In round-trip tests, F2Asm reassembles each CUBIN's disassembled SASS, and all compared executable text sections match the originals exactly. Joint training with F2Asm yields one shared encoder for five Blackwell SM targets and another for three Rubin SM targets, providing strong evidence of a common SASS encoding scheme for instructions shared within each family. Continual training extends the Rubin encoder to all 17,159 previously unsupported cuTile and GROMACS queries with 1,504 additional basis rows, matching the derived lower bound.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-24 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.20532 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T10:27:55.929Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).