Design Flow for GPU and Multicore Execution of Dynamic Dataflow Programs

Abstract

Dataflow programming has received increasing attention in the age of multicore and heterogeneous computing. Modular and concurrent dataflow program descriptions enable highly automated approaches for design space exploration, optimization and deployment of applications. A great advance in dataflow programming has been the recent introduction of the RVC-CAL language. Having been standardized by the ISO, the RVC-CAL dataflow language provides a solid basis for the development of tools, design methodologies and design flows. This paper proposes a novel design flow for mapping RVC-CAL dataflow programs to parallel and heterogeneous execution platforms. Through the proposed design flow the programmer can describe an application in the RVC-CAL language and map it to multi-and many-core platforms, as well as GPUs, for efficient execution. The functionality and efficiency of the proposed approach is demonstrated by a parallel implementation of a video processing application and a run-time reconfigurable filter for telecommunications. Experiments are performed on GPU and multicore platforms with up to 16 cores, and the results show that for high-performance applications the proposed design flow provides up to 4× higher throughput than the state-ofthe-art approach in multicore execution of RVC-CAL programs.

Dynamic SIMD Parallel Execution on GPU from High-Level Dataflow Synthesis

“…Increasing attention has been paid over the past few years to dataflow programming in the field of programming multicore and heterogeneous computing platforms [17][18][19][20][21][22]. As outlined by the various studies that can be found in the literature, one of the most prominent use cases is the execution of dataflow programs on GPU platforms.…”

“…As outlined by the various studies that can be found in the literature, one of the most prominent use cases is the execution of dataflow programs on GPU platforms. In this direction, refer for example to [17][18][19][20][21][22]. Specific to dynamic programs and implemented on GPU platforms using RVC-CAL, the works presented in [23] and in [24] show two different approaches on how to use OpenCL APIs to implement, partition, and subsequently execute RVC-CAL SW programs.…”

Tracing and Profiling Machine Learning Dataflow Applications on GPU

“…It is a data-centric model and can be represented as a graph, where nodes are the operations that are applied to the incoming data, and edges represent the flow of data. It is used in different domains like signal processing [2], [3], [4], [5] and image [6] or video processing [7]. New languages based on this approach have also been developed [8], [9], [10] and [11].…”