FLARE: A Dataflow-Aware and Scalable Hardware Architecture for Neural-Hybrid Scientific Lossy Compression

TL;DR

FLARE is a novel hardware architecture designed to improve neural-hybrid scientific lossy compression by reducing data access and increasing scalability, significantly boosting throughput and energy efficiency in HPC systems.

Contribution

It introduces a dataflow-aware, scalable hardware architecture for neural-hybrid lossy compression, addressing integration challenges and enhancing performance and energy efficiency.

Findings

Achieves up to 96.07x speedup in runtime.

Improves energy efficiency up to 520.68x.

Reduces off-chip data access and bubble overhead.

Abstract

Scientific simulation leveraging high-performance computing (HPC) systems is crucial for modeling complex systems and phenomena in fields such as astrophysics, climate science, and fluid dynamics, generating massive datasets that often reach petabyte to exabyte scales. However, managing these vast data volumes introduces significant I/O and network bottlenecks, limiting practical performance and scalability. While cutting-edge lossy compression frameworks powered by deep neural networks (DNNs) have demonstrated superior compression ratios by capturing complex data correlations, their integration into HPC workflows poses substantial challenges due to the hybrid non-neural and neural computation patterns, causing excessive memory access overhead, large sequential stalls, and limited adaptability to varying data sizes and workloads in existing hardware platforms. To overcome these challenges and push the limit of high-performance scientific computing, we for the first time propose FLARE, a dataflow-aware and scalable hardware architecture for neural-hybrid scientific lossy compression. FLARE minimizes off-chip data access, reduces bubble overhead through efficient dataflow, and adopts a modular design that provides both scalability and flexibility, significantly enhancing throughput and energy efficiency on modern HPC systems. Particularly, the proposed FLARE achieves runtime speedups ranging from $3.50 \times$ to $96.07 \times$, and energy efficiency improvements ranging from $24.51 \times$ to $520.68 \times$, across various datasets and hardware platforms.