Altimesh Hybridizer Essentials
Hybridizer Essentials is a compiler targeting CUDA-enabled GPUS from .Net. Using parallelization patterns, such as Parallel.For, or ditributing parallel work by hand, the user can benefit from the compute power of GPUS or CPUs without entering the learning curve of CUDA s, all within Visual Studio.
Warning
Installer has changed. Uninstall previous version before installing this one (you may need to use this microsoft tool
License system has changed you need to clear your license and subscription and ask for a new one. Good news : it's now completely free.
Release notes
See more retailed release notes here.
Disclaimer
We don't support the entire C# language or .Net Framework. Main known limitations are
Heap allocation (new A()) from device thread (except for arrays)
System.Collection is not supported
virtual generic functions are not supported
string type is only partially supported (they must be const)
Known Issues
SharedMatrix and cublas samples can fail on windows 8.1 for unknown reason.
NPP_ImageSegmentation fails because NPP library dropped support for some functions in version 12 of the library
Debugging and profiling
If debug information is available, the debugging can be performed from within the development environment, while running the optimized code on the target hardware; for example, a program written in C#, with a breakpoint in the cs file within Visual Studio can have that breakpoint hit and its local variables and object data explored while execution occurs on an NVIDIA GPU (see snapshot).
Customization
While providing automated default behavior, Hybridizer leaves full control to the developer at each phase, allowing the reuse of existing device-specific code, use of existing external libraries or custom handmade code snippets.
Performance
All of this flexibility does not come at the expense of performances loss. As our benchmark illustrate, code generated by the Hybridizer can perform as well as hand-written code.