Solving $k$-Nearest Neighbor Problem on Multiple Graphics Processors

TL;DR

This paper presents a GPU-based algorithm for efficiently solving the $k$-nearest neighbor problem, crucial in recommendation systems, demonstrating significant speedup and scalability on multiple GPUs.

Contribution

The paper introduces a novel GPU algorithm combining an $N$-body problem approach and a partial sort optimized for small $k$, achieving high bandwidth utilization and scalability.

Findings

Over 330x faster on two GPUs compared to CPU implementation.

Effective scaling with increasing number of GPUs.

High bandwidth utilization through coalesced memory access.

Abstract

The recommendation system is a software system to predict customers' unknown preferences from known preferences. In the recommendation system, customers' preferences are encoded into vectors, and finding the nearest vectors to each vector is an essential part. This vector-searching part of the problem is called a $k$-nearest neighbor problem. We give an effective algorithm to solve this problem on multiple graphics processor units (GPUs). Our algorithm consists of two parts: an $N$-body problem and a partial sort. For a algorithm of the $N$-body problem, we applied the idea of a known algorithm for the $N$-body problem in physics, although another trick is need to overcome the problem of small sized shared memory. For the partial sort, we give a novel GPU algorithm which is effective for small $k$. In our partial sort algorithm, a heap is accessed in parallel by threads with a low cost of synchronization. Both of these two parts of our algorithm utilize maximal power of coalesced memory access, so that a full bandwidth is achieved. By an experiment, we show that when the size of the problem is large, an implementation of the algorithm on two GPUs runs more than 330 times faster than a single core implementation on a latest CPU. We also show that our algorithm scales well with respect to the number of GPUs.