arXiv subjects

Evelyne Ringoot

Publications and source records attributed to Evelyne Ringoot.

7 recordsLinked to original sources

arXiv2026

Portable to Efficient: Auto-Tuning Hardware-Agnostic GPU Kernels in Julia

Traditionally, GPU kernels have been developed and optimized within vendor-specific programming models to achieve high performance, resulting in software that is difficult to optimize and adapt across increasingly heterogeneous computing systems. Hardware-agnostic programming models offer a more sustainable approach to GPU software development by improving portability and maintainability, but achieving efficient execution across diverse architectures remains challenging. We address this challenge by integrating auto-tuning into hardware-agnostic GPU kernels written in Julia. We rebuild the established Kernel Tuner auto-tuning framework with Julia support, enabling systematic exploration of kernel configurations for hardware-agnostic GPU kernels targeting NVIDIA, AMD, Intel, and Apple GPUs. We demonstrate this approach on hardware-agnostic singular value decomposition (SVD) as implemented in the NextLA.jl linear algebra library. The results show that auto-tuning is essential for creating resource-efficient hardware-agnostic GPU kernels across a variety of hardware. Optimal configurations improve kernel performance by a factor of 3x to 7x compared to median parameter configurations, demonstrating the substantial impact of tuning on efficient hardware utilization.

cs.PF↗

arXiv2026

Cross-Model Cross-Language AI Coding Agent Performance: Accuracy and Speed of Parallel CLRS Algorithms

AI coding agents have quickly become omnipresent in software engineering. Their serial performance, both in terms of accuracy and speed, has been extensively covered. However, recent initial results suggest their parallel programming capabilities lag behind serial programming capabilities. This paper presents a cross-language evaluation of three coding agents -- Cursor's Composer 2.0, GPT 5.4, and Claude Sonnet 4.6 -- on parallel code generation across three algorithm categories -- sorting, graph traversal, and search -- in C++, Python, and Julia. For each algorithm and language pair, we prompt a coding agent to produce a parallel implementation from a serial baseline, track the prompting effort required to achieve both functional correctness and performance improvements, and measure speedup against both custom serial baselines and third-party library implementations. We find that coding agents can produce correct parallel implementations with modest prompting effort, but that achieving meaningful speedup is heavily algorithm- and language-dependent. Sonnet 4.6 delivers the strongest overall performance gains, whereas GPT 5.4 produces no measurable speedups despite consistent correctness. C++ is most consistently parallelizable for graph algorithms, while Python and Julia achieve the largest speedups on search algorithms: no single language dominates across all categories. Python and Julia each achieve speedup on some graph algorithms but regress on others. These findings underscore the impact of including runtime performance efficiency as a main LLM performance metric, in addition to accuracy, particularly for parallel implementations.

cs.SE↗

arXiv2026

Accelerating Bidiagonalization of Banded Matrices through Memory-Aware Bulge-Chasing on GPUs

The reduction of a banded matrix to bidiagonal form is a critical step in the calculation of Singular Values, a cornerstone of scientific computing and AI. Although inherently parallel, this step has traditionally been considered unsuitable for GPUs due to its memory-bound nature. However, recent advances in GPU architectures, such as increased L1 memory per Streaming Multiprocessor or Compute Unit and larger L2 caches, have shifted this paradigm. In this work, we present the first GPU-accelerated algorithm for reducing a banded matrix to bidiagonal form, integrated into an open-source software package. Our algorithm builds on prior multicore CPU cache-efficient bulge-chasing methods, adapted to modern GPU architectures to optimize throughput. Leveraging Julia's high-level array abstractions and KernelAbstractions.jl, we implement a single function that is both hardware-agnostic and data-precision-aware, running efficiently across NVIDIA, AMD, Intel, and Apple Metal GPUs. We develop a hardware-aware performance model to guide tuning and identify key hyperparameters that govern optimal GPU performance for memory-bound workloads. We show that such workloads, when carefully optimized, can achieve substantial speed-ups on modern GPUs: our implementation outperforms multithreaded CPU libraries (PLASMA,SLATE) starting from matrix sizes as small as 1024x1024, and achieves over 100x speed-up on 32k x 32k matrices. Moreover, the algorithm's performance scales linearly with the matrix bandwidth, enabling efficient reduction of matrices with larger bandwidths, previously considered impractical.

cs.DC↗

arXiv2026

Hierarchical Recursive Precision for Accelerating Symmetric Linear Solves on MXUs

Symmetric positive-definite system solvers based on Cholesky factorization are fundamental to many scientific applications, such as climate modeling. We present a portable, nested recursive mixed-precision solver designed for Matrix Processing Units (MXUs), including NVIDIA Tensor Cores (H200) and AMD Matrix Cores (MI300X), that assigns low-precision FP16 arithmetic to large off-diagonal blocks, while preserving high precision on diagonal blocks to ensure numerical stability. The solver is implemented in Julia, providing a high-level, hardware-agnostic interface. We demonstrate up to a 5.07x speedup relative to the diagonal-precision vendor baseline, with 100x better accuracy than pure half precision on H200, providing higher accuracy than low-precision at higher speed than high-precision. Positive performance trends are also observed on MI300X, demonstrating broad applicability across GPUs.

cs.DC↗

arXiv2025

Performant Unified GPU Kernels for Portable Singular Value Computation Across Hardware and Precision

This paper presents a portable, GPU-accelerated implementation of a QR-based singular value computation algorithm in Julia. The singular value ecomposition (SVD) is a fundamental numerical tool in scientific computing and machine learning, providing optimal low-rank matrix approximations. Its importance has increased even more in large-scale machine learning pipelines, including large language models (LLMs), where it enables low-rank adaptation (LoRA). The implemented algorithm is based on the classic two-stage QR reduction, consisting of successive matrix reduction to band form and bidiagonal form. Our implementation leverages Julia's multiple dispatch and metaprogramming capabilities, integrating with the GPUArrays and KernelAbstractions frameworks to provide a unified type and hardware-agnostic function. It supports diverse GPU architectures and data types, and is, to our knowledge, the first GPU-accelerated singular value implementation to support Apple Metal GPUs and half precision. Performance results on multiple GPU backends and data types demonstrate that portability does not require sacrificing performance: the unified function outperforms most linear algebra libraries (MAGMA, SLATE, rocSOLVER, oneMKL) for matrix sizes larger than 1024x1024, and achieves 80%-90% of the performance of cuSOLVER for large matrices.

cs.DC↗

arXiv2025

Toward Portable GPU Performance: Julia Recursive Implementation of TRMM and TRSM

This paper presents a performant and portable recursive implementation of triangular matrix-matrix multiplication (TRMM) and triangular solve (TRSM) in Julia for GPUs, two kernels that underlie many linear-algebra algorithms. We restructure TRMM and TRSM so that most work is executed as general matrix-matrix multiplication (GEMM), improving use of the GPU memory hierarchy and reducing latency. Exploiting Julia's multiple dispatch and metaprogramming together with the GPUArrays and KernelAbstractions frameworks, we expose a single hardware-agnostic API that runs on NVIDIA, AMD, and Apple Silicon GPUs. For large matrices the recursive code reaches throughput comparable to vendor libraries such as cuBLAS and rocBLAS, while providing these routines on Apple Silicon for the first time. The entire implementation is only a few hundred lines of code, showing that unified Julia programs can deliver near-vendor performance across heterogeneous architectures.

cs.MS↗

arXiv2021

Stick-slip phenomena and Schallamach waves captured using reversible cohesive elements

Reversibility is of paramount importance in the correct representation of surface peeling in various physical settings, ranging from motility in nature, to gripping devices in robotic applications, and even to sliding of tectonic plates. Modeling the detachment-reattachment sequence, known as stick-slip, imposes several challenges in a continuum framework. Here we exploit customized reversible cohesive elements in a hybrid finite element model that can handle occurrence of snap-through instabilities. The simulations capture various peeling phenomena that emerge in experimental observations, where layers are pulled from a flat, rigid substrate in the direction parallel to the surface. For long layers, periodicity in reattachment is shown to develop and is linked to the concept of Schallamach waves. Further, the connection between surface properties and stick-slip behavior is investigated: we find that stick-slip is linked to the propensity of the interface to localize deformation and damage. Beyond elucidating the various peeling behaviors and the detachment modes, the computational framework developed here provides a straightforward approach for investigation of complex delamination processes, which can guide development of future applications across different scales and in various settings.

cond-mat.soft↗