paper-with-me

Papers

QiMeng-Xpiler: Transcompiling Tensor Programs for Deep Learning Systems with a Neural-Symbolic Approach

2025-05-04 · Shouyang Dong, Yuanbo Wen, Jun Bi, Di Huang, Jiaming Guo, Jianxing Xu, Ruibai Xu, Xinkai Song, Yifan Hao, Xuehai Zhou, Tianshi Chen, Qi Guo, Yunji Chen

Heterogeneous deep learning systems (DLS) such as GPUs and ASICs have been widely deployed in industrial data centers, which requires to develop multiple low-level tensor programs for different platforms. An attractive solution to relieve the programming burden is to transcompile the legacy code of one platform to others. However, current transcompilation techniques struggle with either tremendous manual efforts or functional incorrectness, rendering "Write Once, Run Anywhere" of tensor programs an open question. We propose a novel transcompiler, i.e., QiMeng-Xpiler, for automatically translating tensor programs across DLS via both large language models (LLMs) and symbolic program synthesis, i.e., neural-symbolic synthesis. The key insight is leveraging the powerful code generation ability of LLM to make costly search-based symbolic synthesis computationally tractable. Concretely, we propose multiple LLM-assisted compilation passes via pre-defined meta-prompts for program transformation. During each program transformation, efficient symbolic program synthesis is employed to repair incorrect code snippets with a limited scale. To attain high performance, we propose a hierarchical auto-tuning approach to systematically explore both the parameters and sequences of transformation passes. Experiments on 4 DLS with distinct programming interfaces, i.e., Intel DL Boost with VNNI, NVIDIA GPU with CUDA, AMD MI with HIP, and Cambricon MLU with BANG, demonstrate that QiMeng-Xpiler correctly translates different tensor programs at the accuracy of 95% on average, and the performance of translated programs achieves up to 2.0x over vendor-provided manually-optimized libraries. As a result, the programming productivity of DLS is improved by up to 96.0x via transcompiling legacy tensor programs.

📄 PDF Abstract BibTeX arXiv:2505.02146

Code (0)

등록된 구현이 없습니다.

Tasks

Code GenerationGPUProgram Synthesis

Similar Papers 제목 키워드 기반

QiMeng: Fully Automated Hardware and Software Design for Processor Chip

2025-06-05 · Rui Zhang, Yuanbo Wen, Shuyao Cheng, Di Huang 외

Processor chip design technology serves as a key frontier driving breakthroughs in computer science and related fields. With the rapid advancement of information technology, conventional design paradigms face three major…

TensorFI: A Flexible Fault Injection Framework for TensorFlow Applications

2020-04-03 · Zitao Chen, Niranjhana Narayanan, Bo Fang, Guanpeng Li 외

As machine learning (ML) has seen increasing adoption in safety-critical domains (e.g., autonomous vehicles), the reliability of ML systems has also grown in importance. While prior studies have proposed techniques to en…

Autonomous Vehicles

QiMeng-SALV: Signal-Aware Learning for Verilog Code Generation

2025-10-22 · Yang Zhang, Rui Zhang, Jiaming Guo, Lei Huang 외 arxiv

The remarkable progress of Large Language Models (LLMs) presents promising opportunities for Verilog code generation which is significantly important for automated circuit design. The lacking of meaningful functional rew…

Reinforcement LearningCode Generation

Ansor: Generating High-Performance Tensor Programs for Deep Learning

2020-06-11 · Lianmin Zheng, Chengfan Jia, Minmin Sun, Zhao Wu 외

High-performance tensor programs are crucial to guarantee efficient execution of deep neural networks. However, obtaining performant tensor programs for different operators on various hardware platforms is notoriously ch…

CPUDeep LearningGPUVocal Bursts Intensity Prediction

Learning to Optimize Tensor Programs

2018-05-21 · NeurIPS 2018 12 · Tianqi Chen, Lianmin Zheng, Eddie Yan, Ziheng Jiang 외

We introduce a learning-based framework to optimize tensor programs for deep learning workloads. Efficient implementations of tensor operators, such as matrix multiplication and high dimensional convolution, are key enab…

CPUDeep LearningGPU