Quantum algorithms which accept hot qubit inputs

Realistic physical implementations of quantum computers can entail tradeoffs which depart from the ideal model of quantum computation. Although these tradeoffs have allowed successful demonstration of certain quantum algorithms, a crucial question is whether they fundamentally limit the computational capacity of such machines. We study the limitations of a quantum computation model in which only ensemble averages of measurement observables are accessible. Furthermore, we stipulate that input qubits may only be prepared in highly random, ``hot'' mixed states. In general, these limitations are believed to dramatically detract from the computational power of the system. However, we construct a class of algorithms for this limited model, which, surprisingly, are polynomially equivalent to the ideal case. This class includes the well known Deutsch-Jozsa algorithm.