Graphics card firm Nvidia has announced the release of its NVPerfKit 2.1, a collection of performance tools for the debugging and profiling of OpenGL and Direct3D softwar...
Graphics card firm Nvidia has announced the release of its NVPerfKit 2.1, a collection of performance tools for the debugging and profiling of OpenGL and Direct3D software applications for Windows and Linux. Developers working on 32- and 64-bit Linux or Windows platforms can utilize the same technology that drives the Nvidia NVPerfHUD performance analysis tool, a profiling and visual debugging heads-up display (HUD) program for Direct3D applications supporting Microsoft DirectX 9.0c and Windows XP on select Nvidia graphics processing units. According to Nvidia representatives, developers using NVPerfHUD 4 experience an average performance improvement of 35%, and find an average of 11 rendering bugs. Representatives also note that by offering access to low-level performance counters inside the driver, as well as hardware counters inside the GPU, NVPerfKit 2.1 helps developers determine how applications use the GPU by identifying and confirming resolved performance issues for substantial time savings during application development. Among the features included within NVPerfKit 2.1 are an instrumented driver that interfaces with the graphics API and GPU to provide performance counter data; the NVPerfHUD, which offers real-time analysis of Direct3D applications; and NVPerfSDK, which includes an API for accessing GPU signals in applications, sample code and helper classes for OpenGL and DirectX applications, and an Nvidia Developer Control Panel and instrumented driver. The release also includes a Nvidia plug in for Microsoft PIX for Windows, GLExpert for debugging of OpenGL usage errors and performance issues, and support for PerfMon, Intel VTune, gDEBugger and other analysis tools. The NVPerfKit 2.1 tool suite, which can be downloaded from the Nvidia website, supports Windows XP and Linux on Nvidia GeForce 7 Series GPUs, Nvidia GeForce 6 Series GPUs, and Nvidia Quadro FX or more recent Nvidia GPUs. In addition, Nvidia notes that older GPUs are supported with reduced functionality.
You May Also Like
