Viktor Kuncak, Eva Darulova
Modern computing has adopted the floating point type as a default way to describe computations with real numbers. Thanks to dedicated hardware support, such computations are efficient on modern architectures. However, rigorous reasoning about the resulting ...
2011