26 - 04 - 2024
Login Form



 


Share this post

Submit to FacebookSubmit to TwitterSubmit to LinkedIn

AMD (NYSE: AMD) today announced a significant update to its CodeXL unified tool suite for developers designed to accelerate applications running on AMD processors.  CodeXL is AMD’s primary developer tool for generating detailed information about the real-world performance of AMD Central Processing Units (CPUs), Accelerated Processing Units (APUs) and Graphics Processing Units (GPUs). AMD CodeXL version 1.6 provides GPU profiling capabilities and API-level debugging for OpenCL™ 2.0 applications. CodeXL 1.6 also introduces a new advanced tool for profiling APU power consumption, which enables developers to reduce the power requirements of their application. These features significantly bolster the tools available to developers for creating applications on either Windows® or Linux® that take advantage of the inherent processing capability of heterogeneous system architecture (HSA).

The CodeXL tool suite assists software developers and ISVs to utilize parallel programming by harnessing the compute power of AMD’s high-performance CPUs, GPUs and APUs. CodeXL 1.6 includes advanced GPU debugging, GPU and CPU profiling and static OpenCL kernel analysis. With AMD CodeXL 1.6, developers can bring to life applications with tremendous performance, long battery life, and smooth playback. AMD has also released a beta version of the AMD APP SDK 3.0 with comprehensive OpenCL 2.0 support. AMD APP SDK 3.0 contains a complete set of sample code illustrating how to use each of the major new features of OpenCL 2.0.

“With the release of the CodeXL 1.6 tool suite and AMD APP SDK 3.0, developers gain greater capability to unlock the power of HSA hardware features such as shared memory through heterogeneous Unified Memory Architecture,” said Manju Hegde, AMD corporate vice president, Heterogeneous Applications and Solutions. “OpenCL 2.0 is a state-of–the-art parallel programming platform and CodeXL1.6 provides extensive debugging and performance profiling capabilities to enable software developers to more readily take advantage of heterogeneous systems.”

The new power profiler samples a variety of counters that capture individual APU sub-components, including their power consumption, effective frequency and temperature. The profile data is displayed in real time using a rich and intuitive graphic representation, while the application session is in progress. A command line option is also available.