NVIDIA CUDA TOOLKIT COMPATIBILITY UPDATE
NVIDIA CUDA TOOLKIT COMPATIBILITY DRIVER

For example, an application built against the CUDA 3.2 SDK will continue to function even on today’s driver stack. The CUDA driver (libcuda.so) provides binary backward compatibility. We define binary compatibility as a set of guarantees provided by the library, where an application targeting the said library will continue to work when dynamically linked against a different version of the library. Although the driver APIs can change, they are versioned, and their symbols persist across releases to maintain binary compatibility.

Developers are notified through deprecation and documentation mechanisms of any current or upcoming changes. APIs can be deprecated and removed, requiring changes to the application. We define source compatibility as a set of guarantees provided by the library, where a well-formed application built against a specific version of the library (using the SDK) will continue to build and run without errors when a newer version of the SDK is installed.īoth the CUDA driver and the CUDA runtime are not source compatible across the different SDK releases. This allows the use of newer toolkits on existing system installations, providing improvements and features of the latest CUDA while minimizing the risks associated with new driver deployments. Starting with CUDA 10.0, NVIDIA introduced a new forward-compatible upgrade path that allows the kernel mode components on the system to remain untouched, while the CUDA driver is upgraded. the user mode components (the CUDA driver, the OpenGL driver, etc.).the kernel mode components (the ‘display’ driver), and.The driver components are further separated into two categories: the development libraries (including the CUDA runtime), and.The CUDA toolkit consists of two main components:

The current upgrade path for CUDA is shown in Figure 1. The new CUDA Compatibility Platform enables the use of new CUDA toolkit components on systems with older base installations.

NVIDIA CUDA TOOLKIT COMPATIBILITY UPDATE

However, the tight coupling of the CUDA runtime with the display driver (specifically libcuda.so-the CUDA driver on Linux systems), means that the customer is required to update the entire driver stack to use the latest CUDA software (including the compiler, libraries, and tools). The CUDA toolkit is transitioning to a faster release cadence to deliver new features, performance improvements, and critical bug fixes. CUDA Compatibility document describes the use of new CUDA toolkit components on systems with older base installations.

Nvidia cuda toolkit compatibility