SparsePipe: Parallel Deep Learning for 3D Point Clouds
This work addresses the memory and performance bottlenecks in training deep learning models on 3D point clouds, which is a problem for researchers and practitioners working with 3D data.
This paper introduces SparsePipe, an asynchronous parallelism approach for multi-GPU training of 3D point clouds using sparse tensor representations and generalized convolutions. It significantly reduces memory requirements and improves performance on point cloud benchmarks compared to dense solutions.
We propose SparsePipe, an efficient and asynchronous parallelism approach for handling 3D point clouds with multi-GPU training. SparsePipe is built to support 3D sparse data such as point clouds. It achieves this by adopting generalized convolutions with sparse tensor representation to build expressive high-dimensional convolutional neural networks. Compared to dense solutions, the new models can efficiently process irregular point clouds without densely sliding over the entire space, significantly reducing the memory requirements and allowing higher resolutions of the underlying 3D volumes for better performance. SparsePipe exploits intra-batch parallelism that partitions input data into multiple processors and further improves the training throughput with inter-batch pipelining to overlap communication and computing. Besides, it suitably partitions the model when the GPUs are heterogeneous such that the computing is load-balanced with reduced communication overhead. Using experimental results on an eight-GPU platform, we show that SparsePipe can parallelize effectively and obtain better performance on current point cloud benchmarks for both training and inference, compared to its dense solutions.