LTTng CLUST: A System-Wide Unified CPU and GPU Tracing Tool for OpenCL Applications. (18th August 2015)

• Record Type: Journal Article
• Authors: Couturier, David; Dagenais, Michel R.
• Other Names: Marzolla Moreno Academic Editor.
• Is Part Of: Advances in software engineering. Volume 2015(2015)
• Publication Date: 2015-08-18
• Subjects: Software engineering -- Periodicals; Software engineering; Periodicals; Electronic journals; 005.1
• DOI: 10.1155/2015/940628
• Languages: English
• ISSNs: 1687-8655

Abstract

Abstract : As computation schemes evolve and many new tools become available to programmers to enhance the performance of their applications, many programmers started to look towards highly parallel platforms such as Graphical Processing Unit (GPU). Offloading computations that can take advantage of the architecture of the GPU is a technique that has proven fruitful in recent years. This technology enhances the speed and responsiveness of applications. Also, as a side effect, it reduces the power requirements for those applications and therefore extends portable devices battery life and helps computing clusters to run more power efficiently. Many performance analysis tools such asLTTng, strace andSystemTap already allow Central Processing Unit (CPU) tracing and help programmers to use CPU resources more efficiently. On the GPU side, different tools such asNvidia 'sNsight, AMD 'sCodeXL, and third partyTAU andVampirTrace allow tracing Application Programming Interface (API) calls and OpenCL kernel execution. These tools are useful but are completely separate, and none of them allow a unified CPU-GPU tracing experience. We propose an extension to the existing scalable and highly efficientLTTng tracing platform to allow unified tracing of GPU along with CPU's full tracing capabilities.