EngineCL: Usability and Performance in Heterogeneous Computing

Abstract

Heterogeneous systems have become one of the most common architectures today, thanks to their excellent performance and energy consumption. However, due to their heterogeneity they are very complex to program and even more to achieve performance portability on different devices. This paper presents EngineCL, a new OpenCL-based runtime system that outstandingly simplifies the co-execution of a single massive data-parallel kernel on all the devices of a heterogeneous system. It performs a set of low level tasks regarding the management of devices, their disjoint memory spaces and scheduling the workload between the system devices while providing a layered API. EngineCL has been validated in two compute nodes (HPC and commodity system), that combine six devices with different architectures. Experimental results show that it has excellent usability compared with OpenCL; a maximum 2.8% of overhead compared to the native version under loads of less than a second of execution and a tendency towards zero for longer execution times; and it can reach an average efficiency of 0.89 when balancing the load.

Exploiting co-execution with oneAPI: heterogeneity from a modern perspective

“…However, it is well known that static scheduling cannot adapt to the irregular behavior of many applications, leading to significant load imbalances [16]. In these applications, the processing time of a data set depends not only on its size, but also on the nature of the data.…”

“…However, hardware heterogeneity complicates the development of efficient and portable software, especially when specialized components require their own programming models. In this context, some of the hot topics being researched are: supporting single source programming, improving the usability and efficiency of memory space, distributing computation and data among different devices, and load balancing [3,[16][17][18][19][20][21]25].…”

“…This article addresses a new challenge in improving the usability and exploitation of heterogeneous systems, providing oneAPI with the capacity for coexecution. This is defined as the collaboration of all the devices in the system (including the CPU) to execute a single massively data-parallel kernel [15,16,20,25]. However, it is a hard task for the programmer and needs to be done effortless in order to be widely used.…”

Cooperative CPU, GPU, and FPGA heterogeneous execution with EngineCL

“…EngineCL is a runtime acting as an OpenCL C++ wrapper to simplify the programming of heterogeneous devices and squeeze their performance out [21]. It is specially designed to be used in large data-parallel applications and provides three load balancing algorithms.…”

“…EngineCL is a high-level framework that provides scheduling and data management primitives on top of OpenCL, easing the programmability of heterogeneous systems [21]. While EngineCL has been successful in systems coupling a CPU with either a GPU or Xeon Phi device, it does not yet provide support for FPGAs.…”