Vortex: Overcoming Memory Capacity Limitations in GPU-Accelerated Large-Scale Data Analytics

TL;DR

Vortex is a GPU framework that overcomes memory limitations in large-scale data analytics by optimizing IO across multiple GPUs, improving performance and price efficiency without caching data in GPU memory.

Contribution

It introduces an optimized multi-GPU IO primitive, a new programming model separating kernel and IO scheduling, and novel query operators for large-scale data analytics.

Findings

Vortex outperforms Proteus by 5.7× in performance.

Vortex achieves 2.5× better price performance over CPU baseline.

The framework effectively handles data exceeding GPU memory capacity.

Abstract

Despite the high computational throughput of GPUs, limited memory capacity and bandwidth-limited CPU-GPU communication via PCIe links remain significant bottlenecks for accelerating large-scale data analytics workloads. This paper introduces Vortex, a GPU-accelerated framework designed for data analytics workloads that exceed GPU memory capacity. A key aspect of our framework is an optimized IO primitive that leverages all available PCIe links in multi-GPU systems for the IO demand of a single target GPU. It routes data through other GPUs to such target GPU that handles IO-intensive analytics tasks. This approach is advantageous when other GPUs are occupied with compute-bound workloads, such as popular AI applications that typically underutilize IO resources. We also introduce a novel programming model that separates GPU kernel development from IO scheduling, reducing programmer burden and enabling GPU code reuse. Additionally, we present the design of certain important query operators and discuss a late materialization technique based on GPU's zero-copy memory access. Without caching any data in GPU memory, Vortex improves the performance of the state-of-the-art GPU baseline, Proteus, by 5.7$\times$ on average and enhances price performance by 2.5$\times$ compared to a CPU-based DuckDB baseline.