There exist other toolchains that target CUDA GPUs, my comments don’t apply to those toolchains or efforts. My comments above apply to the NVIDIA-provided toolchain, of course. The fact that it does not exist today indicates so far that the cost/benefit scenario is not compelling.
#Mingw cygwin software
But there are definitely resource limitations, and support for a new compiler platform on windows would require a significant expenditure of resource and effort. Installing the 32-bit and 64-bit C (gcc) and C++ (g++) Compilers on Windows (These are different than the compilers included with Cygwin) MinGW (Minimalist GNU for Windows), formerly mingw32, is a free and open source software development environment for creating Microsoft Windows applications. There are no technical issues that absolutely prevent the support of additional compilers. This is so that limited resources can be focused where they will provide the most benefit. If there is enough evidence of need and demand, I’m sure NVIDIA would respond, however a constant theme internally is to look for ways to reduce the support matrix, not increase it.
The situation is not likely to change in the near future.
In order to support other compilers, there would have to be significant market demand, as evidenced by actual usage for large industrial/commercial/enterprise codebases. So it will be the first choice of supported compilers on windows. Large commercial codebases get developed on this platform. Cygwin helps users to port Linux applications to Windows. The motto of Cygwin, which is ‘Get that Linux feeling on Windows’, clearly tells us its intended purpose. The premier compiler on Windows is the one that is provided by the Windows vendor, namely Microsoft. Since both Cygwin and MinGW can be used to execute C programs on Windows, let’s look at the differences between the two and why we ought to choose MinGW over Cygwin. Adding an additional configuration adds an additional dimension to the QA matrix, so the cost to add one major new dimension (like a compiler supported) can be considerable in terms of effort to validate the interface. If you want to see all the moving parts in the CUDA compilation process, run nvcc with the -verbose switch.Īny interface must be validated. So there is actual technical effort to create the interface in the first place.
Also, object file formats must be completely compatible. For example data structure alignment in memory must match, for every possible/conceivable arrangement of data types and qualifiers. There are at least 2 general reasons, maybe 3ĬUDA uses a fairly tight integration with the host compiler.
0 kommentar(er)
