# Libra: Unleashing GPU Heterogeneity for High-Performance Sparse Matrix Multiplication

arXiv: https://arxiv.org/abs/2506.22714 · HTML: https://mlsanity.com/paper/2506.22714/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 57 references that are in the map.
- First posted on arXiv 28 Jun 2025.

## Builds on (most cited first, 20 of 20)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Inductive Representation Learning on Large Graphs](https://mlsanity.com/paper/1706.02216/) (2017)
- [How Powerful are Graph Neural Networks?](https://mlsanity.com/paper/1810.00826/) (2018)
- [Sparse GPU Kernels for Deep Learning](https://mlsanity.com/paper/2006.10901/) (2020)
- [Attention-based Graph Neural Network for Semi-supervised Learning](https://mlsanity.com/paper/1803.03735/) (2018)
- [SparseTIR: Composable Abstractions for Sparse Compilation in Deep Learning](https://mlsanity.com/paper/2207.04606/) (2022)
- [GE-SpMM: General-purpose Sparse Matrix-Matrix Multiplication on GPUs for Graph Neural Networks](https://mlsanity.com/paper/2007.03179/) (2020)
- [Learning Graph Representations with Embedding Propagation](https://mlsanity.com/paper/1710.03059/) (2017)
- [IGB: Addressing The Gaps In Labeling, Features, Heterogeneity, and Size of Public Graph Datasets for Deep Learning Research](https://mlsanity.com/paper/2302.13522/) (2023)
- [QGTC: Accelerating Quantized Graph Neural Networks via GPU Tensor Core](https://mlsanity.com/paper/2111.09547/) (2021)
- [TC-GNN: Bridging Sparse GNN Computation and Dense Tensor Cores on GPUs](https://mlsanity.com/paper/2112.02052/) (2021)
- [Efficient Quantized Sparse Matrix Operations on Tensor Cores](https://mlsanity.com/paper/2209.06979/) (2022)
- [Tag-Aware Ordinal Sparse Factor Analysis for Learning and Content Analytics](https://mlsanity.com/paper/1412.5967/) (2014)
- [High Performance Unstructured SpMM Computation Using Tensor Cores](https://mlsanity.com/paper/2408.11551/) (2024)
- [Cached Operator Reordering: A Unified View for Fast GNN Training](https://mlsanity.com/paper/2308.12093/) (2023)
- [UniSparse: An Intermediate Language for General Sparse Format Customization](https://mlsanity.com/paper/2403.05802/) (2024)
- [Acc-SpMM: Accelerating General-purpose Sparse Matrix-Matrix Multiplication with GPU Tensor Cores](https://mlsanity.com/paper/2501.09251/) (2025)
- [FlashSparse: Minimizing Computation Redundancy for Fast Sparse Matrix Multiplications on Tensor Cores](https://mlsanity.com/paper/2412.11007/) (2024)
- [HC-SpMM: Accelerating Sparse Matrix-Matrix Multiplication for Graphs with Hybrid GPU Cores](https://mlsanity.com/paper/2412.08902/) (2024)

## Built on by (most cited first, 1 of 1)

- [LocalBench: Benchmarking LLMs on County-Level Local Knowledge and Reasoning](https://mlsanity.com/paper/2511.10459/) (2025)

## Abstract (arXiv)

Sparse matrix multiplication operators (i.e., SpMM and SDDMM) are widely used in deep learning and scientific computing. Modern accelerators are commonly equipped with Tensor Core Units (TCUs) and CUDA cores to accelerate sparse operators. The former excels at structured matrix computations, whereas the latter offers greater programming flexibility. However, how to combine these two resources to maximize sparse-operator performance remains unclear. In this work, we first identify the source of performance gains in hybrid computation and systematically analyze their complementary strengths. Motivated by this, we propose Libra, a holistic framework that efficiently leverages heterogeneous computing resources to accelerate both SpMM and SDDMM operators. Specifically, Libra introduces a 2D-aware (locality and utilization) workload distribution method to precisely identify the optimal task mapping, simultaneously leveraging the data reuse capabilities of TCUs and the flexibility of CUDA cores to minimize computational redundancy. Libra further incorporates hybrid load balancing, occupancy-aware task scheduling, and efficient kernel implementations to maximize execution efficiency. Extensive experiments on H100 and RTX 4090 GPUs demonstrate that Libra surpasses all the 12 up-to-date baselines significantly, e.g., on average 1.77x speedup over FlashSparse, 1.73x over RoDe, and 2.9x over DGL for end-to-end GNN applications. Libra opens up a new perspective for sparse operator acceleration by fully unleashing the power of heterogeneous GPU resources.