DAG-based Scheduling with Resource Sharing for Multi-task Applications in a Polyglot GPU Runtime

Abstract

GPUs are readily available in cloud computing and personal devices, but their use for data processing acceleration has been slowed down by their limited integration with common programming languages such as Python or Java. Moreover, using GPUs to their full capabilities requires expert knowledge of asynchronous programming. In this work, we present a novel GPU run time scheduler for multi-task GPU computations that transparently provides asynchronous execution, space-sharing, and transfer-computation overlap without requiring in advance any information about the program dependency structure. We leverage the GrCUDA polyglot API to integrate our scheduler with multiple high-level languages and provide a platform for fast prototyping and easy GPU acceleration. We validate our work on 6 benchmarks created to evaluate task-parallelism and show an average of 44% speedup against synchronous execution, with no execution time slowdown compared to hand-optimized host code written using the C++ CUDA Graphs API.

AStitch: enabling a new multi-dimensional optimization space for memory-intensive ML training and inference on modern SIMT architectures

“…CUDA Graph [5] binds, but not fuses, GPU kernels to reduce kernel launch overhead, which still suffers from off-chip memory traffic. Furthermore, it results in high GPU memory consumption to store all the graph metadata of every kernel [35]. AStitch does not have these problems and explores a larger optimization scope beyond CUDA Graph.…”