Memory access profiling tools for alpha-based architectures

The development of efficient algorithms on today’s high performance computers is far from straight-forward. Applications need to take full advantage of the deep memory hierarchy which implies that the user has to know exactly how his/her implementation gets executed. With today’s compilers it can be very difficult to understand or predict the execution path without having to look at the machine code. We present a set of tools designed to help us better understand programs and their memory access pattern. These tools enable the programmer to compare memory access patterns of different algorithms as well as provide insight into the algorithm’s behavior e.g. potential bottlenecks resulting from memory accesses.