Exploring the features of OpenCL 2.0

Mukherjee, Saoni; Gong, Xiang; Yu, Liang; McCardwell, Carter; Ukidave, Yash; Dao, Tuan; Paravecino, Fanny Nina; Kaeli, David

doi:10.1145/2791321.2791326

Cited by 5 publications

(1 citation statement)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors believe that UMA can lead to performance improvements in benchmarks where a small subset of the data is repeatedly accessed. In addition, Mukherjee et al 36 compared the performance of OpenCL 2.0 and OpenCL 1.2 versions of four benchmarks (AES, FIR, IIR, and HMM) on an AMD R9 290X GPU. One of the features they used in the OpenCL 2.0 versions was SVM coarse‐grained buffers.…”

Section: Related Workmentioning

confidence: 99%

Compilation of MATLAB computations to CPU/GPU via C/OpenCL generation

Reis

Bispo

Cardoso

2020

Concurrency and Computation

View full text Add to dashboard Cite

In order to take advantage of the processing power of current computing platforms, programmers typically need to develop software versions for different target devices. This task is time-consuming and requires significant programming and computer architecture expertise. A possible and more convenient alternative is to start with a single high-level description of a program with minimum implementation details, and generate custom implementations according to the target platform. In this paper, we use MATLAB as a high-level programming language and propose a compiler that targets CPU/GPU computing platforms by generating customized implementations in C and OpenCL. We propose a number of compiler techniques to automatically generate efficient C and OpenCL code from MATLAB programs. One of such compiler techniques relies on heuristics to decide when and how to use Shared Virtual Memory (SVM). The experimental results show that our approach is able to generate code that provides significant speedups (eg, geometric mean speedup of 11× for a set of simple benchmarks) using a discrete GPU over equivalent sequential C code executing on a CPU. With more complex benchmarks, for which only some code regions can be parallelized, and are thus offloaded, the generated code achieved speedups of up to 2.2×. We also show the impact of using SVM, specifically fine-grained buffers, and the results show that the compiler is able to achieve significant speedups, both over the versions without SVM and with naïve aggressive SVM use, across three CPU/GPU platforms.

show abstract

Section: Related Workmentioning

confidence: 99%