I do believe we are just starting to understand how to efficiently code for multiple processers/threads. IBM and M$ are working on languages specifically to fix these problems. Actually, I think IBM's work is based on C++, but it will help tons since the future of increasing speed is in adding more cores.







