With libcu++, Nvidia is not saying how or when software should be written. Because the library is meeting the C++ standard, it does exactly what you said you wa

With libcu++, Nvidia is not saying how or when software should be written. Because the library is meeting the C++ standard, it does exactly what you said you want, it allows any developer to write code for any GPU (or CPU!) The library is doing the thing you‘re asking for. AMD & Intel can support the same code with only namespace changes, using their own version, because it’s open and written to the open standard.