Skip to content

Learning Exact NVIDIA SASS Encoders with $\mathbb{F}_2$ Linear Algebra

Aug 2026 · 0 citations · 24 references
Computer Science

TL;DR

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 and uses Gaussian elimination over $\mathbb{F}_2$ to incrementally build a compact basis, detect inconsistencies, and reject inputs outside the learned span.

Abstract

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.

View source

Similar papers

Precision As You Need: Stochastic Computing Is a Dense Adaptive Quantizer

This work builds a GPU library that emulates SC matrix multiplication at scale, exposes stream lengths as first-class kernel arguments, and evaluates SC end-to-end on image classification, object detection and instance segmentation, class-conditional image generation, and visual world-model planning.

Hao-Ran Jin, Kang-Qi Zhang, Ji-Rong Yang et al. · 0 citations
#artificial intelligence Preprint Oct 2026

Cleave: Scaling Tensor Program Optimization via Decoupled Algebraic Search and Operator Scheduling

Optimized kernels such as FlashAttention and FlashDecoding are crucial for accelerating today's large models. Most of them are handwritten by experts because existing ML compilers cannot match their efficiency. Producing such kernels requires fusing computations with multiple reductions, which requires both algebraic t...

David Pissarra, Jinkun Lin, Hai-Tian Jiang et al. · 0 citations
#artificial intelligence Preprint Oct 2026

Cascadia: Resident 975B MoE Inference on Eleven AI PCs

This work presents Cascadia's resident execution of Inkling, a 975B-total/41B-active-parameter model, and contributes a custom resident MoE engine that preserves Inkling's routing rules, constructs compressed graphs for OpenVINO's fused iGPU primitives, and coordinates FP16 expert computation with FP32 output restorati...

Tate Berenbaum, Matias Parij, M. Venkatachalam · 0 citations

Accelerating the Mitigation of LLM Inference Nondeterminism Across GPU Architectures

This work presents a solution employing a set of fixed-configuration fused-upcast GEMM kernels that load 16-bit weights from memory, upcast them to FP32 in registers, and accumulate with IEEE-754 arithmetic in a reduction order that is a pure function of the problem shape and is therefore independent of the device, its...

Liam Cooper, Shinnung Jeong, Hyeran Jeon et al. · 1 citation
Preprint Oct 2026

Integer-Exact Microscaling: Running MXFP4 Language Models on INT8 Matrix Units

This work shows when no per-block work is needed: a row of MXFP4 weights is exactly one INT8 vector times one power-of-two scale if its block exponents span at most three, so a linear layer becomes a single INT8 GEMM with one rescale per output.

Shinyoung Kang, Junhyung Ahn, Yujun Kim et al. · 0 citations
Preprint Oct 2026

OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization

NVFP4 is an attractive datatype for large language model (LLM) inference, offering compact storage and native tensor-core acceleration. However, preserving accuracy using NVFP4 requires careful quantization. In this work we develop a novel quantization scheme called Optimized Smoothing and Scaling for NVFP4 (OSFP4). Fo...

N. B. Ben-David, Or Meir, Or Ordentlich · 0 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

Microsoft Research Blog Nov 16, 2026

TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient

AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.