Lune

HPDC2024Top-tier venue

CereSZ: Enabling and Scaling Error-bounded Lossy Compression on Cerebras CS-2

Shihui Song, Yafan Huang, Peng Jiang, Xiaodong Yu, Weijian Zheng, Sheng Di, Qinglei Cao, Yunhe Feng, Zhen Xie, Franck Cappello

Abstract

Today's scientific applications running on supercomputers produce large volumes of data, leading to critical data storage and communication challenges. To tackle the challenges, error-bounded lossy compression is commonly adopted since it can reduce data size drastically within a user-defined error threshold. Previous work has shown that compression techniques can significantly reduce the storage and I/O overhead while retaining good data quality. However, the existing compressors are mainly designed for CPU and GPU. As new AI chips are being incorporated into supercomputers and increasingly used for accelerating scientific computing, there is a growing demand for efficient data compression on the new architecture. In this paper, we propose an efficient lossy compressor, CereSZ, based on the Cerebras CS-2 system. The compression algorithm is mapped onto Cerebras using both data parallelism and pipeline parallelism. In order to achieve a balanced workload on each processing unit, we propose an algorithm to evenly distribute the pipeline stages. Our experiments with six scientific datasets demonstrate that CereSZ can achieve a throughput from 227.93 GB/s to 773.8 GB/s, 2.43x to 10.98x faster than existing GPU compressors.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

2 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Time-Varying Vector Field Compression with Preserved Critical Point Trajectories cites Optimizing Error-Bounded Lossy Compression for Scientific Data by Dynamic Spline Interpolation
  • Time-Varying Vector Field Compression with Preserved Critical Point Trajectories cites cuSZp: An Ultra-fast GPU Error-bounded Lossy Compression Framework with Optimized End-to-End Performance
  • Optimizing Error-Bounded Lossy Compression for Scientific Data by Dynamic Spline Interpolation cites Significantly Improving Lossy Compression for HPC Datasets with Second-Order Prediction and Parameter Optimization
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers2

  1. Time-Varying Vector Field Compression with Preserved Critical Point TrajectoriesMingze Xia, Yuxiao Li, Pu Jiao, Bei Wang et al.ICDE 2026
  2. POLAR-PIC: A Holistic Framework for Matrixized PIC with Co-Designed Compute, Layout, and CommunicationYizhuo Rao, Xingjian Cui, Shangzhi Pang, Jiabin Xie et al.HPDC 2026

Builds on10

  1. Optimizing Error-Bounded Lossy Compression for Scientific Data by Dynamic Spline InterpolationKai Zhao, Sheng Di, Maxim Dmitriev, Thierry-Laurent D. Tonellot et al.ICDE 2021 · 151 citations
  2. Significantly Improving Lossy Compression for HPC Datasets with Second-Order Prediction and Parameter OptimizationKai Zhao, Sheng Di, Xin Liang, Sihuan Li et al.HPDC 2020 · 77 citations
  3. cuSZp: An Ultra-fast GPU Error-bounded Lossy Compression Framework with Optimized End-to-End PerformanceYafan Huang, Sheng Di, Xiaodong Yu, Guanpeng Li et al.SC 2023 · 52 citations
  4. Ultrafast Error-bounded Lossy Compression for Scientific DatasetsXiaodong Yu, Sheng Di, Kai Zhao, Jiannan Tian et al.HPDC 2022 · 41 citations
  5. COMET: A Novel Memory-Efficient Deep Learning Training Framework by Using Error-Bounded Lossy CompressionSian Jin, Chengming Zhang, Xintong Jiang, Yunhe Feng et al.VLDB 2022 · 39 citations

Related papers

  1. FZ-GPU: A Fast and High-Ratio Lossy Compressor for Scientific Computing Applications on GPUsBoyuan Zhang, Jiannan Tian, Sheng Di, Xiaodong Yu et al.HPDC 2023 · 27 citations
  2. Efficient Lossless Compression of Scientific Floating-Point Data on CPUs and GPUsNoushin Azami, Alex Fallin, Martin BurtscherASPLOS 2025 · 18 citations
  3. cuSZ-i: High-Ratio Scientific Lossy Compression on GPUs with Optimized Multi-Level InterpolationJinyang Liu, Jiannan Tian, Shixun Wu, Sheng Di et al.SC 2024 · 17 citations
  4. Boosting Scientific Error-Bounded Lossy Compression through Optimized Synergistic Lossy-Lossless OrchestrationShixun Wu, Jinwen Pan, Jinyang Liu, Jiannan Tian et al.SC 2025 · 6 citations
  5. waveSZ: a hardware-algorithm co-design of efficient lossy compression for scientific dataJiannan Tian, Sheng Di, Chengming Zhang, Xin Liang et al.PPoPP 2020 · 26 citations
Dusk over the sea between two cliffs drawn in fine vertical lines