FT K-means: A High-Performance K-means on GPU with Fault Tolerance

TL;DR

FT K-means is a GPU-accelerated clustering algorithm that significantly improves performance and introduces fault tolerance, effectively handling soft errors with minimal overhead on modern NVIDIA GPUs.

Contribution

The paper presents a novel high-performance GPU implementation of K-means with online fault tolerance and a warp-level tensor-core error correction scheme.

Findings

FT K-means outperforms cuML's implementation by 10-300% on irregular data shapes.

Fault tolerance incurs only 11% overhead, maintaining robustness under error injection.

The approach achieves competitive performance with enhanced resilience to soft errors.

Abstract

K-means is a widely used algorithm in clustering, however, its efficiency is primarily constrained by the computational cost of distance computing. Existing implementations suffer from suboptimal utilization of computational units and lack resilience against soft errors. To address these challenges, we introduce FT K-means, a high-performance GPU-accelerated implementation of K-means with online fault tolerance. We first present a stepwise optimization strategy that achieves competitive performance compared to NVIDIA's cuML library. We further improve FT K-means with a template-based code generation framework that supports different data types and adapts to different input shapes. A novel warp-level tensor-core error correction scheme is proposed to address the failure of existing fault tolerance methods due to memory asynchronization during copy operations. Our experimental evaluations on NVIDIA T4 GPU and A100 GPU demonstrate that FT K-means without fault tolerance outperforms cuML's K-means implementation, showing a performance increase of 10\%-300\% in scenarios involving irregular data shapes. Moreover, the fault tolerance feature of FT K-means introduces only an overhead of 11\%, maintaining robust performance even with tens of errors injected per second.